Skip to content

Instantly share code, notes, and snippets.

@mpuig

mpuig/Intent Classification with Embeddings.ipynb

Created September 29, 2024 17:46
Show Gist options
  • Save mpuig/0eb18599ab79691cf593cccddd06d137 to your computer and use it in GitHub Desktop.
Save mpuig/0eb18599ab79691cf593cccddd06d137 to your computer and use it in GitHub Desktop.
Download ZIP
Intent Classification with Embeddings
Display the source blob
Display the rendered blob
Raw
Intent Classification with Embeddings.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 21,
"id": "3df59c49-8233-4ab9-9e6c-0d40212317b4",
"metadata": {},
"outputs": [],
"source": [
"!pip install transformers torch einops --quiet\n",
"!pip install 'numpy<2' --quiet\n",
"!pip install tqdm pandas numpy scikit-learn --quiet\n",
"!pip install keras tensorflow --quiet\n",
"!pip install pyarrow matplotlib seaborn --quiet"
]
},
{
"cell_type": "code",
"id": "b84a7199-6141-40a3-bdca-00cadc38b362",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:07:37.065408Z",
"start_time": "2024-09-29T15:07:37.034518Z"
}
},
"source": [
"import json\n",
"import re\n",
"import time\n",
"from ast import literal_eval\n",
"from pathlib import Path\n",
"\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"import pandas as pd\n",
"from keras import layers, models, optimizers, losses, callbacks\n",
"from sklearn.decomposition import PCA\n",
"from sklearn.ensemble import RandomForestClassifier\n",
"from sklearn.manifold import TSNE\n",
"from sklearn.metrics import classification_report, accuracy_score\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.preprocessing import LabelEncoder\n",
"from sklearn.utils import shuffle\n",
"from tqdm.auto import tqdm\n",
"from transformers import AutoModel"
],
"outputs": [],
"execution_count": 2
},
{
"cell_type": "code",
"id": "684a2137-90a2-421f-9958-1cdab278e083",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:07:46.325868Z",
"start_time": "2024-09-29T15:07:43.654929Z"
}
},
"source": [
"df = pd.read_parquet(\n",
" \"hf://datasets/urvog/llama2_transcripts_healthcare_callcenter/data/train-00000-of-00001-643ef6d7d804c522.parquet\")\n",
"df.head()"
],
"outputs": [
{
"data": {
"text/plain": [
" text\n",
"0 <s>[INST] Classify the following call transcri...\n",
"1 <s>[INST] Classify the following call transcri...\n",
"2 <s>[INST] Classify the following call transcri...\n",
"3 <s>[INST] Classify the following call transcri...\n",
"4 <s>[INST] Classify the following call transcri..."
],
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>text</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"execution_count": 3
},
{
"cell_type": "code",
"id": "03b1df3a-f8cf-44b3-af7a-a20757e0303c",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:10:09.397663Z",
"start_time": "2024-09-29T15:10:09.395700Z"
}
},
"source": [
"text = str(df.iloc[0].text).replace(\"\\n\", \"\")\n",
"print(f\"{text[:100]} [...]{text[-100:]}\")"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"<s>[INST] Classify the following call transcript:Agent 3: Thank you for calling HealthHarbor, my nam [...] Have a great day!Agent 3: You too, Emma. Goodbye! [/INST] Medical Advice or Symptom Assessment </s>\n"
]
}
],
"execution_count": 13
},
{
"cell_type": "code",
"id": "8ddfb317-01cd-4e7f-8d41-ebf0b2f6b979",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:11:31.400073Z",
"start_time": "2024-09-29T15:11:31.270932Z"
}
},
"source": [
"def clean_and_extract_text(text):\n",
" text = re.sub(r'<s>\\[INST\\] Classify the following call transcript:\\n\\n', '', text)\n",
" match = re.search(r'\\[/INST\\] (.+?) </s>', text)\n",
" if match:\n",
" intent = match.group(1)\n",
" text = text[:match.start()].strip()\n",
" else:\n",
" intent = None\n",
" return text, intent\n",
"\n",
"\n",
"df[['cleaned_text', 'intent']] = df['text'].apply(lambda x: pd.Series(clean_and_extract_text(x)))"
],
"outputs": [],
"execution_count": 17
},
{
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:12:07.431865Z",
"start_time": "2024-09-29T15:12:07.428880Z"
}
},
"cell_type": "code",
"source": [
"# Initialize and apply LabelEncoder\n",
"label_encoder = LabelEncoder()\n",
"df['intent_encoded'] = label_encoder.fit_transform(df['intent'])\n",
"\n",
"intent_mapping = dict(zip(label_encoder.classes_, label_encoder.transform(label_encoder.classes_)))\n",
"print(\"Intent mapping:\", intent_mapping)"
],
"id": "67b8eca4282be7aa",
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Intent mapping: {'Appointment Scheduling': 0, 'Follow-up Calls and Care Management': 1, 'General Inquiries': 2, 'Healthcare Facility Information': 3, 'Insurance and Billing Inquiries': 4, 'Lab and Test Results': 5, 'Medical Advice or Symptom Assessment': 6, 'Medical Records Requests': 7, 'Medication Refills and Prescription Inquiries': 8, 'Referrals and Specialist Appointments': 9}\n"
]
}
],
"execution_count": 20
},
{
"cell_type": "code",
"id": "1e9d1950-2043-4217-8165-46be19104ce2",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:12:31.215581Z",
"start_time": "2024-09-29T15:12:31.211442Z"
}
},
"source": [
"# Check the distribution of intents in the dataset\n",
"intent_distribution = df['intent'].value_counts().reset_index()\n",
"\n",
"print(\"Intent Distribution:\")\n",
"print(intent_distribution)"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Intent Distribution:\n",
" intent count\n",
"0 Insurance and Billing Inquiries 112\n",
"1 Medical Records Requests 112\n",
"2 Healthcare Facility Information 111\n",
"3 Medical Advice or Symptom Assessment 102\n",
"4 Follow-up Calls and Care Management 100\n",
"5 Lab and Test Results 97\n",
"6 Medication Refills and Prescription Inquiries 95\n",
"7 Referrals and Specialist Appointments 95\n",
"8 General Inquiries 89\n",
"9 Appointment Scheduling 87\n"
]
}
],
"execution_count": 21
},
{
"cell_type": "code",
"id": "a0c73b81-a27e-427e-b59b-1f8733ca8a49",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:13:10.842025Z",
"start_time": "2024-09-29T15:13:10.833712Z"
}
},
"source": [
"def balance_dataset(df):\n",
" min_count = df['intent'].value_counts().min()\n",
" balanced_df = df.groupby('intent').apply(lambda x: x.sample(min_count)).reset_index(drop=True)\n",
" return balanced_df\n",
"\n",
"\n",
"balanced_df = balance_dataset(df)\n",
"\n",
"# Shuffle the balanced dataset\n",
"balanced_df = shuffle(balanced_df, random_state=42)"
],
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/var/folders/r3/rx5lk2194sqd33p9mkdv1xg80000gn/T/ipykernel_68588/2732155417.py:3: DeprecationWarning: DataFrameGroupBy.apply operated on the grouping columns. This behavior is deprecated, and in a future version of pandas the grouping columns will be excluded from the operation. Either pass `include_groups=False` to exclude the groupings or explicitly select the grouping columns after groupby to silence this warning.\n",
" balanced_df = df.groupby('intent').apply(lambda x: x.sample(min_count)).reset_index(drop=True)\n"
]
}
],
"execution_count": 22
},
{
"cell_type": "code",
"id": "7537aae7-df26-4352-9757-cc671ceac345",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:13:31.087850Z",
"start_time": "2024-09-29T15:13:31.081155Z"
}
},
"source": [
"# Split the data into training and temporary (which will be further split into validation and test)\n",
"train_df, temp_df = train_test_split(balanced_df, test_size=0.3, stratify=balanced_df['intent'], random_state=42)\n",
"\n",
"# Further split the temporary dataset into validation and test sets\n",
"validation_df, test_df = train_test_split(temp_df, test_size=0.5, stratify=temp_df['intent'], random_state=42)\n",
"\n",
"print(f\"Training set: {len(train_df)} samples\")\n",
"print(f\"Validation set: {len(validation_df)} samples\")\n",
"print(f\"Test set: {len(test_df)} samples\")"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Training set: 609 samples\n",
"Validation set: 130 samples\n",
"Test set: 131 samples\n"
]
}
],
"execution_count": 23
},
{
"cell_type": "code",
"id": "d768d064-90c6-406a-b7dd-29ff0f2513f8",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:13:58.121451Z",
"start_time": "2024-09-29T15:13:58.090125Z"
}
},
"source": [
"# save the balanced dataframes\n",
"train_df.to_parquet('data/balanced_train_dataset.parquet', index=False)\n",
"validation_df.to_parquet('data/balanced_validation_dataset.parquet', index=False)\n",
"test_df.to_parquet('data/balanced_test_dataset.parquet', index=False)"
],
"outputs": [],
"execution_count": 24
},
{
"cell_type": "code",
"id": "6e5d3943-71f4-4c96-8e3a-dfd4887e5da3",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:15:04.550353Z",
"start_time": "2024-09-29T15:15:04.544070Z"
}
},
"source": [
"# Extract intent counts from each dataset\n",
"train_counts = train_df['intent'].value_counts()\n",
"validation_counts = validation_df['intent'].value_counts()\n",
"test_counts = test_df['intent'].value_counts()\n",
"\n",
"# Create a DataFrame to print a table with the distribution data\n",
"distribution_df = pd.DataFrame({\n",
" 'Intent': train_counts.index,\n",
" 'Training Count': train_counts.values,\n",
" 'Validation Count': validation_counts.reindex(train_counts.index, fill_value=0).values,\n",
" 'Test Count': test_counts.reindex(train_counts.index, fill_value=0).values\n",
"})\n",
"distribution_df.head(20)"
],
"outputs": [
{
"data": {
"text/plain": [
" Intent Training Count \\\n",
"0 Medication Refills and Prescription Inquiries 61 \n",
"1 Lab and Test Results 61 \n",
"2 Appointment Scheduling 61 \n",
"3 Healthcare Facility Information 61 \n",
"4 Follow-up Calls and Care Management 61 \n",
"5 Referrals and Specialist Appointments 61 \n",
"6 Insurance and Billing Inquiries 61 \n",
"7 Medical Records Requests 61 \n",
"8 General Inquiries 61 \n",
"9 Medical Advice or Symptom Assessment 60 \n",
"\n",
" Validation Count Test Count \n",
"0 13 13 \n",
"1 13 13 \n",
"2 13 13 \n",
"3 13 13 \n",
"4 13 13 \n",
"5 13 13 \n",
"6 13 13 \n",
"7 13 13 \n",
"8 13 13 \n",
"9 13 14 "
],
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>Intent</th>\n",
" <th>Training Count</th>\n",
" <th>Validation Count</th>\n",
" <th>Test Count</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Medication Refills and Prescription Inquiries</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>Lab and Test Results</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>Appointment Scheduling</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>Healthcare Facility Information</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>Follow-up Calls and Care Management</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>Referrals and Specialist Appointments</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>Insurance and Billing Inquiries</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>Medical Records Requests</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>General Inquiries</td>\n",
" <td>61</td>\n",
" <td>13</td>\n",
" <td>13</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>Medical Advice or Symptom Assessment</td>\n",
" <td>60</td>\n",
" <td>13</td>\n",
" <td>14</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
]
},
"execution_count": 25,
"metadata": {},
"output_type": "execute_result"
}
],
"execution_count": 25
},
{
"cell_type": "code",
"id": "3f7a0863-18bf-4a2c-a050-4c5f67bf255b",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:15:35.539458Z",
"start_time": "2024-09-29T15:15:27.470549Z"
}
},
"source": [
"# Let's start with embeddings\n",
"EMBEDDING_MODEL_NAME = \"jinaai/jina-embeddings-v3\"\n",
"embedding_model = AutoModel.from_pretrained(EMBEDDING_MODEL_NAME, trust_remote_code=True)"
],
"outputs": [],
"execution_count": 26
},
{
"cell_type": "code",
"id": "c597b10a-ff1f-4489-8a29-17cb34db5b9d",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:15:48.109646Z",
"start_time": "2024-09-29T15:15:48.105650Z"
}
},
"source": [
"tqdm.pandas()\n",
"\n",
"\n",
"def embed_text(text: str, model) -> list[float]:\n",
" \"\"\"Generates embeddings for a given text using the provided model.\"\"\"\n",
" # You can customize the task as needed: 'retrieval.query', 'retrieval.passage', 'classification', etc.\n",
" embeddings = model.encode(text, task=\"classification\")\n",
" return embeddings\n",
"\n",
"\n",
"def create_embeddings(model, df):\n",
" df['embeddings'] = df['cleaned_text'].progress_apply(lambda text: embed_text(text, model))\n",
" return df"
],
"outputs": [],
"execution_count": 27
},
{
"cell_type": "code",
"id": "cad0d51a-4f70-4207-a53d-2bc3349cc493",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:17:09.436783Z",
"start_time": "2024-09-29T15:17:07.777464Z"
}
},
"source": [
"DATA_PATH = 'data/'\n",
"\n",
"train_path = Path(DATA_PATH) / 'call_center_train.csv'\n",
"test_path = Path(DATA_PATH) / 'call_center_test.csv'\n",
"\n",
"if not train_path.exists():\n",
" df_train = create_embeddings(embedding_model, train_df)\n",
" df_train['embeddings'] = df_train['embeddings'].apply(lambda emb: json.dumps(emb.tolist()))\n",
" df_train.to_csv(train_path, index=False)\n",
"\n",
"if not test_path.exists():\n",
" df_test = create_embeddings(embedding_model, test_df)\n",
" df_test['embeddings'] = df_test['embeddings'].apply(lambda emb: json.dumps(emb.tolist()))\n",
" df_test.to_csv(test_path, index=False)\n",
"\n",
"df_train = pd.read_csv(train_path)\n",
"df_test = pd.read_csv(test_path)\n",
"\n",
"df_train[\"embeddings\"] = df_train.embeddings.apply(literal_eval).apply(np.array)\n",
"df_test[\"embeddings\"] = df_test.embeddings.apply(literal_eval).apply(np.array)\n",
"\n",
"df_train['intent'] = df_train['intent'].astype('category')\n",
"df_test['intent'] = df_test['intent'].astype('category')\n",
"\n",
"df_train.head()"
],
"outputs": [
{
"data": {
"text/plain": [
" text \\\n",
"0 <s>[INST] Classify the following call transcri... \n",
"1 <s>[INST] Classify the following call transcri... \n",
"2 <s>[INST] Classify the following call transcri... \n",
"3 <s>[INST] Classify the following call transcri... \n",
"4 <s>[INST] Classify the following call transcri... \n",
"\n",
" cleaned_text \\\n",
"0 Agent 3: Thank you for calling HealthHarbor. M... \n",
"1 Agent 1: Thank you for calling HealthHarbor, m... \n",
"2 Agent: Thank you for calling HealthHarbor, my ... \n",
"3 Agent 3: Thank you for calling HealthHarbor, m... \n",
"4 Agent 3: Thank you for calling HealthHarbor, m... \n",
"\n",
" intent \\\n",
"0 Medication Refills and Prescription Inquiries \n",
"1 Medication Refills and Prescription Inquiries \n",
"2 Medical Advice or Symptom Assessment \n",
"3 Lab and Test Results \n",
"4 Medication Refills and Prescription Inquiries \n",
"\n",
" embeddings intent_encoded \n",
"0 [0.06473791599273682, -0.06422188878059387, 0.... 8 \n",
"1 [0.04085281491279602, -0.05341055616736412, 0.... 8 \n",
"2 [0.06634803861379623, -0.11376452445983887, 0.... 6 \n",
"3 [0.08077315986156464, -0.10844098031520844, 0.... 5 \n",
"4 [0.08430707454681396, -0.08557997643947601, 0.... 8 "
],
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>text</th>\n",
" <th>cleaned_text</th>\n",
" <th>intent</th>\n",
" <th>embeddings</th>\n",
" <th>intent_encoded</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" <td>Agent 3: Thank you for calling HealthHarbor. M...</td>\n",
" <td>Medication Refills and Prescription Inquiries</td>\n",
" <td>[0.06473791599273682, -0.06422188878059387, 0....</td>\n",
" <td>8</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" <td>Agent 1: Thank you for calling HealthHarbor, m...</td>\n",
" <td>Medication Refills and Prescription Inquiries</td>\n",
" <td>[0.04085281491279602, -0.05341055616736412, 0....</td>\n",
" <td>8</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" <td>Agent: Thank you for calling HealthHarbor, my ...</td>\n",
" <td>Medical Advice or Symptom Assessment</td>\n",
" <td>[0.06634803861379623, -0.11376452445983887, 0....</td>\n",
" <td>6</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" <td>Agent 3: Thank you for calling HealthHarbor, m...</td>\n",
" <td>Lab and Test Results</td>\n",
" <td>[0.08077315986156464, -0.10844098031520844, 0....</td>\n",
" <td>5</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>&lt;s&gt;[INST] Classify the following call transcri...</td>\n",
" <td>Agent 3: Thank you for calling HealthHarbor, m...</td>\n",
" <td>Medication Refills and Prescription Inquiries</td>\n",
" <td>[0.08430707454681396, -0.08557997643947601, 0....</td>\n",
" <td>8</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
]
},
"execution_count": 28,
"metadata": {},
"output_type": "execute_result"
}
],
"execution_count": 28
},
{
"cell_type": "code",
"id": "019126a9-72ba-4f89-a524-b448cba98200",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:19:11.929641Z",
"start_time": "2024-09-29T15:19:11.914059Z"
}
},
"source": [
"def build_and_compile_nn_classification_model(input_size: int, num_classes: int) -> models.Model:\n",
" inputs = layers.Input(shape=(input_size,))\n",
" x = layers.Dense(128, activation='relu')(inputs)\n",
" x = layers.Dense(num_classes, activation='sigmoid')(x)\n",
"\n",
" model = models.Model(inputs, x)\n",
" model.compile(\n",
" loss=losses.SparseCategoricalCrossentropy(),\n",
" optimizer=optimizers.Adam(learning_rate=0.001),\n",
" metrics=['accuracy']\n",
" )\n",
" return model\n",
"\n",
"\n",
"embedding_size = len(df_train['embeddings'].iloc[0])\n",
"num_classes = len(df_train['intent'].unique())\n",
"nn_classifier = build_and_compile_nn_classification_model(embedding_size, num_classes)\n",
"nn_classifier.summary()"
],
"outputs": [
{
"data": {
"text/plain": [
"\u001B[1mModel: \"functional_1\"\u001B[0m\n"
],
"text/html": [
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"functional_1\"</span>\n",
"</pre>\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃\u001B[1m \u001B[0m\u001B[1mLayer (type) \u001B[0m\u001B[1m \u001B[0m┃\u001B[1m \u001B[0m\u001B[1mOutput Shape \u001B[0m\u001B[1m \u001B[0m┃\u001B[1m \u001B[0m\u001B[1m Param #\u001B[0m\u001B[1m \u001B[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ input_layer_1 (\u001B[38;5;33mInputLayer\u001B[0m) │ (\u001B[38;5;45mNone\u001B[0m, \u001B[38;5;34m1024\u001B[0m) │ \u001B[38;5;34m0\u001B[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_2 (\u001B[38;5;33mDense\u001B[0m) │ (\u001B[38;5;45mNone\u001B[0m, \u001B[38;5;34m128\u001B[0m) │ \u001B[38;5;34m131,200\u001B[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_3 (\u001B[38;5;33mDense\u001B[0m) │ (\u001B[38;5;45mNone\u001B[0m, \u001B[38;5;34m10\u001B[0m) │ \u001B[38;5;34m1,290\u001B[0m │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
],
"text/html": [
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃<span style=\"font-weight: bold\"> Layer (type) </span>┃<span style=\"font-weight: bold\"> Output Shape </span>┃<span style=\"font-weight: bold\"> Param # </span>┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ input_layer_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">InputLayer</span>) │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1024</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>) │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">131,200</span> │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>) │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">1,290</span> │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"</pre>\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"\u001B[1m Total params: \u001B[0m\u001B[38;5;34m132,490\u001B[0m (517.54 KB)\n"
],
"text/html": [
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">132,490</span> (517.54 KB)\n",
"</pre>\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"\u001B[1m Trainable params: \u001B[0m\u001B[38;5;34m132,490\u001B[0m (517.54 KB)\n"
],
"text/html": [
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">132,490</span> (517.54 KB)\n",
"</pre>\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"\u001B[1m Non-trainable params: \u001B[0m\u001B[38;5;34m0\u001B[0m (0.00 B)\n"
],
"text/html": [
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
"</pre>\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"execution_count": 30
},
{
"cell_type": "code",
"id": "ee5ed243-f78a-4aac-967a-0bd995304bcd",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:19:46.638296Z",
"start_time": "2024-09-29T15:19:45.721571Z"
}
},
"source": [
"# Train the neural network classifier\n",
"\n",
"x_train = np.stack(df_train['embeddings'])\n",
"y_train = df_train['intent_encoded']\n",
"\n",
"x_val = np.stack(df_test['embeddings'])\n",
"y_val = df_test['intent_encoded']\n",
"\n",
"start_time = time.time()\n",
"callback = callbacks.EarlyStopping(monitor='accuracy', patience=3)\n",
"history = nn_classifier.fit(\n",
" x=x_train,\n",
" y=y_train,\n",
" validation_data=(x_val, y_val),\n",
" batch_size=32,\n",
" epochs=20,\n",
" callbacks=[callback],\n",
" verbose=0\n",
")\n",
"training_time_nn = time.time() - start_time\n",
"print(f\"Neural Network Training Time: {training_time_nn:.4f} seconds\")"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Neural Network Training Time: 0.8836 seconds\n"
]
}
],
"execution_count": 31
},
{
"cell_type": "code",
"id": "42e1673a-60bd-4f2f-89ec-a1c6c5c7a4f6",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:23:45.033515Z",
"start_time": "2024-09-29T15:23:45.003256Z"
}
},
"source": [
"# Evaluate model performance\n",
"nn_classifier.evaluate(x=x_val, y=y_val, return_dict=True)"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001B[1m5/5\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 875us/step - accuracy: 0.7505 - loss: 1.3184\n"
]
},
{
"data": {
"text/plain": [
"{'accuracy': 0.7633587718009949, 'loss': 1.3109252452850342}"
]
},
"execution_count": 36,
"metadata": {},
"output_type": "execute_result"
}
],
"execution_count": 36
},
{
"cell_type": "code",
"id": "e080c656-84ef-4f48-b54e-5e24f4b2a43e",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:20:12.799881Z",
"start_time": "2024-09-29T15:20:12.741842Z"
}
},
"source": [
"# Evaluate the model's performance using scikit-learn's classification report and accuracy score\n",
"\n",
"# Predict the class probabilities on the validation set\n",
"y_pred_probs = nn_classifier.predict(x_val)\n",
"\n",
"# Use np.argmax to convert the predicted probabilities to class labels (for multi-class classification)\n",
"y_pred = np.argmax(y_pred_probs, axis=1)\n",
"\n",
"accuracy_nn = accuracy_score(y_val, y_pred)\n",
"report_nn = classification_report(y_val, y_pred)\n",
"\n",
"print(\"Neural Network Classifier:\")\n",
"print(f\"Accuracy: {accuracy_nn:.4f}\")\n",
"print(\"Classification Report:\")\n",
"print(report_nn)"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001B[1m5/5\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 3ms/step \n",
"Neural Network Classifier:\n",
"Accuracy: 0.7634\n",
"Classification Report:\n",
" precision recall f1-score support\n",
"\n",
" 0 0.50 0.85 0.63 13\n",
" 1 1.00 0.46 0.63 13\n",
" 2 1.00 0.08 0.14 13\n",
" 3 0.54 1.00 0.70 13\n",
" 4 0.87 1.00 0.93 13\n",
" 5 0.87 1.00 0.93 13\n",
" 6 0.82 1.00 0.90 14\n",
" 7 1.00 1.00 1.00 13\n",
" 8 0.93 1.00 0.96 13\n",
" 9 0.75 0.23 0.35 13\n",
"\n",
" accuracy 0.76 131\n",
" macro avg 0.83 0.76 0.72 131\n",
"weighted avg 0.83 0.76 0.72 131\n",
"\n"
]
}
],
"execution_count": 33
},
{
"cell_type": "code",
"id": "19cd4e3d-b8d0-4d92-a7ec-888d48dc98fe",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:26:11.686128Z",
"start_time": "2024-09-29T15:26:10.900172Z"
}
},
"source": [
"# Random forest classifier\n",
"\n",
"# Set up the training and testing data\n",
"X_train = df_train[\"embeddings\"].to_list() # Embeddings as input features\n",
"y_train = df_train[\"intent_encoded\"].values # Encoded label as target\n",
"\n",
"X_test = df_test[\"embeddings\"].to_list() # Test embeddings as input features\n",
"y_test = df_test[\"intent_encoded\"].values # Test encoded labels as target\n",
"\n",
"# Train Random Forest Classifier\n",
"rf_classifier = RandomForestClassifier(n_estimators=100, random_state=42)\n",
"\n",
"start_time = time.time()\n",
"rf_classifier.fit(X_train, y_train)\n",
"training_time_rf = time.time() - start_time\n",
"print(f\"Random Forest Training Time: {training_time_rf:.4f} seconds\")\n",
"\n",
"# Generate and print classification report\n",
"predictions = rf_classifier.predict(X_test)\n",
"accuracy_rf = accuracy_score(y_test, predictions)\n",
"report_rf = classification_report(y_test, predictions)\n",
"\n",
"print(f\"Accuracy: {accuracy_rf:.4f}\")\n",
"print(report_rf)"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Random Forest Training Time: 0.7730 seconds\n",
"Accuracy: 0.8855\n",
" precision recall f1-score support\n",
"\n",
" 0 0.85 0.85 0.85 13\n",
" 1 0.83 0.77 0.80 13\n",
" 2 0.75 0.46 0.57 13\n",
" 3 0.68 1.00 0.81 13\n",
" 4 1.00 1.00 1.00 13\n",
" 5 0.92 0.92 0.92 13\n",
" 6 1.00 0.93 0.96 14\n",
" 7 1.00 1.00 1.00 13\n",
" 8 1.00 1.00 1.00 13\n",
" 9 0.86 0.92 0.89 13\n",
"\n",
" accuracy 0.89 131\n",
" macro avg 0.89 0.89 0.88 131\n",
"weighted avg 0.89 0.89 0.88 131\n",
"\n"
]
}
],
"execution_count": 37
},
{
"cell_type": "code",
"id": "266ce414-9083-4bab-9c83-ad5d3ffae6a3",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:43:13.900191Z",
"start_time": "2024-09-29T15:43:13.893179Z"
}
},
"source": [
"def calculate_centroids(df):\n",
" \"\"\"\n",
" Calculates the centroid (mean embedding) for each class in the DataFrame.\n",
" Returns a dictionary mapping class names to their centroid embeddings.\n",
" \"\"\"\n",
" centroids = {}\n",
" for class_name, group in df.groupby('intent'):\n",
" embeddings = np.stack(group['embeddings'].values)\n",
" centroid = np.mean(embeddings, axis=0)\n",
" centroids[class_name] = centroid\n",
" return centroids\n",
"\n",
"\n",
"# Calculate centroids from the training data\n",
"start_time = time.time()\n",
"centroids = calculate_centroids(df_train)\n",
"centroid_calculation_time = time.time() - start_time\n",
"print(f\"Calculation Time: {centroid_calculation_time:.4f} seconds\")\n",
"\n",
"# Print centroids for each category\n",
"for class_name, centroid in centroids.items():\n",
" print(f\"Centroid for class {class_name}: {centroid}\")"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Calculation Time: 0.0036 seconds\n",
"Centroid for class Appointment Scheduling: [ 0.07974444 -0.09036183 0.05841044 ... 0.00359572 -0.0431507\n",
" 0.01681495]\n",
"Centroid for class Follow-up Calls and Care Management: [ 0.07706775 -0.10277535 0.05199573 ... 0.00434913 -0.03911032\n",
" 0.02410924]\n",
"Centroid for class General Inquiries: [ 0.07575651 -0.10408848 0.05828649 ... 0.00125108 -0.04294898\n",
" 0.01620166]\n",
"Centroid for class Healthcare Facility Information: [ 0.07921933 -0.10978018 0.061943 ... 0.00075152 -0.04204226\n",
" 0.01693048]\n",
"Centroid for class Insurance and Billing Inquiries: [ 0.05502919 -0.07886479 0.04231495 ... 0.00089979 -0.0390809\n",
" 0.01733674]\n",
"Centroid for class Lab and Test Results: [ 0.0714327 -0.09749463 0.05169902 ... -0.00197828 -0.03334681\n",
" 0.02281804]\n",
"Centroid for class Medical Advice or Symptom Assessment: [ 0.06933453 -0.10076225 0.04846012 ... 0.00825433 -0.04249502\n",
" 0.00523547]\n",
"Centroid for class Medical Records Requests: [ 0.08010617 -0.11073439 0.06356257 ... -0.01089169 -0.05086842\n",
" 0.02593634]\n",
"Centroid for class Medication Refills and Prescription Inquiries: [ 0.06178747 -0.0644517 0.04638839 ... -0.0004582 -0.04714779\n",
" 0.01738786]\n",
"Centroid for class Referrals and Specialist Appointments: [ 0.07184907 -0.09645086 0.05094772 ... 0.00150695 -0.03803058\n",
" 0.01892385]\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/var/folders/r3/rx5lk2194sqd33p9mkdv1xg80000gn/T/ipykernel_68588/2762160429.py:7: FutureWarning: The default of observed=False is deprecated and will be changed to True in a future version of pandas. Pass observed=False to retain current behavior or observed=True to adopt the future default and silence this warning.\n",
" for class_name, group in df.groupby('intent'):\n"
]
}
],
"execution_count": 44
},
{
"cell_type": "code",
"id": "5fc2e4ea-0b94-4e06-a27d-4401d2aa956d",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:43:17.332709Z",
"start_time": "2024-09-29T15:43:17.300592Z"
}
},
"source": [
"def predict_centroid(X_test, centroids):\n",
" \"\"\"\n",
" Predicts classes for test samples using centroid-based classification.\n",
"\n",
" Args:\n",
" X_test (np.array): Test embeddings.\n",
" centroids (dict): Class centroids.\n",
"\n",
" Returns:\n",
" np.array: Predicted class indices.\n",
" \"\"\"\n",
" centroid_matrix = np.vstack(list(centroids.values()))\n",
" predictions = []\n",
"\n",
" for embedding in X_test:\n",
" embedding = embedding.reshape(1, -1)\n",
" similarities = cosine_similarity(embedding, centroid_matrix)[0]\n",
" best_index = np.argmax(similarities)\n",
" predictions.append(best_index)\n",
"\n",
" return np.array(predictions)\n",
"\n",
"\n",
"start_time = time.time()\n",
"y_pred_centroid_indices = predict_centroid(X_test, centroids)\n",
"prediction_time_centroid = time.time() - start_time\n",
"\n",
"class_names_list = list(centroids.keys())\n",
"predicted_class_names = [class_names_list[idx] for idx in y_pred_centroid_indices]\n",
"predicted_encoded_labels = [df_test['intent'].cat.categories.get_loc(name) for name in predicted_class_names]\n",
"\n",
"# Compute evaluation metrics\n",
"accuracy_centroid = accuracy_score(y_test, predicted_encoded_labels)\n",
"report_centroid = classification_report(y_test, predicted_encoded_labels, target_names=df_test['intent'].cat.categories)\n",
"\n",
"print(\"Centroid-Based Classifier:\")\n",
"print(f\"Prediction Time: {prediction_time_centroid:.4f} seconds\")\n",
"print(f\"Accuracy: {accuracy_centroid:.4f}\")\n",
"print(\"Classification Report:\")\n",
"print(report_centroid)"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Centroid-Based Classifier:\n",
"Prediction Time: 0.0250 seconds\n",
"Accuracy: 0.8550\n",
"Classification Report:\n",
" precision recall f1-score support\n",
"\n",
" Appointment Scheduling 0.92 0.85 0.88 13\n",
" Follow-up Calls and Care Management 0.90 0.69 0.78 13\n",
" General Inquiries 0.50 0.23 0.32 13\n",
" Healthcare Facility Information 0.62 1.00 0.76 13\n",
" Insurance and Billing Inquiries 0.93 1.00 0.96 13\n",
" Lab and Test Results 0.86 0.92 0.89 13\n",
" Medical Advice or Symptom Assessment 0.88 1.00 0.93 14\n",
" Medical Records Requests 1.00 1.00 1.00 13\n",
"Medication Refills and Prescription Inquiries 0.93 1.00 0.96 13\n",
" Referrals and Specialist Appointments 1.00 0.85 0.92 13\n",
"\n",
" accuracy 0.85 131\n",
" macro avg 0.85 0.85 0.84 131\n",
" weighted avg 0.85 0.85 0.84 131\n",
"\n"
]
}
],
"execution_count": 45
},
{
"cell_type": "code",
"id": "77cd83cc-c968-4685-942c-c9c072a8e8be",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:43:20.036757Z",
"start_time": "2024-09-29T15:43:20.030493Z"
}
},
"source": [
"test_data = validation_df[['cleaned_text', 'intent_encoded', 'intent']].to_dict(orient='records')\n",
"\n",
"class_name_mapping = df_train['intent'].cat.categories\n",
"\n",
"\n",
"def predict_class_name_centroid(text, centroids, model):\n",
" \"\"\" Classifies a sentence based on the closest centroid.\"\"\"\n",
" embedding = embed_text(text, model)\n",
"\n",
" # Stack centroids into a matrix\n",
" centroid_matrix = np.vstack(list(centroids.values()))\n",
" class_names = list(centroids.keys())\n",
"\n",
" # Calculate cosine similarity between the embedding and all centroids\n",
" embedding = embedding.reshape(1, -1) # Reshape embedding to (1, embedding_dim) for matrix multiplication\n",
" similarities = cosine_similarity(embedding, centroid_matrix)[0] # Get similarities for all centroids\n",
"\n",
" # Find the index of the centroid with the highest similarity\n",
" best_index = np.argmax(similarities)\n",
"\n",
" # Get the predicted class name\n",
" predicted_class = class_names[best_index]\n",
" return predicted_class\n",
" \n",
"def predict_class_name_nn(text, model, classifier):\n",
" embedding = embed_text(text, model)\n",
" predicted_output = classifier.predict(embedding.reshape(1, -1))\n",
" predicted_class = np.argmax(predicted_output, axis=1)\n",
" predicted_class_name = class_name_mapping[predicted_class[0]]\n",
" return predicted_class_name\n",
"\n",
"\n",
"def predict_class_name_rf(text, model, classifier):\n",
" embedding = embed_text(text, model)\n",
" predicted_class = classifier.predict(embedding.reshape(1, -1))\n",
" predicted_class_name = class_name_mapping[predicted_class[0]]\n",
" return predicted_class_name\n"
],
"outputs": [],
"execution_count": 46
},
{
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:52:09.456717Z",
"start_time": "2024-09-29T15:43:21.074246Z"
}
},
"cell_type": "code",
"source": [
"import pandas as pd\n",
"from sklearn.metrics.pairwise import cosine_similarity\n",
"\n",
"# Initialize an empty list to collect results\n",
"results = []\n",
"\n",
"# Define a DataFrame for the results\n",
"columns = ['Text', 'Actual Intent', 'Predicted NN', 'Predicted RF', 'Predicted Centroid', 'Correct NN', 'Correct RF',\n",
" 'Correct Centroid']\n",
"results_df = pd.DataFrame(columns=columns)\n",
"\n",
"# Process each entry in the test data\n",
"for entry in test_data:\n",
" text = entry[\"cleaned_text\"]\n",
" actual_intent = entry[\"intent\"]\n",
"\n",
" # Predictions\n",
" predicted_class_name_nn = predict_class_name_nn(text, embedding_model, nn_classifier)\n",
" predicted_class_name_rf = predict_class_name_rf(text, embedding_model, rf_classifier)\n",
" predicted_class_name_centroid = predict_class_name_centroid(text, centroids, embedding_model)\n",
"\n",
" # Determine correctness\n",
" correct_nn = '✅' if actual_intent == predicted_class_name_nn else '❌'\n",
" correct_rf = '✅' if actual_intent == predicted_class_name_rf else '❌'\n",
" correct_centroid = '✅' if actual_intent == predicted_class_name_centroid else '❌'\n",
"\n",
" # Append the result to the list\n",
" results.append({\n",
" 'Text': text[:100], # Show only first 100 characters\n",
" 'Actual Intent': actual_intent,\n",
" 'Predicted NN': predicted_class_name_nn,\n",
" 'Predicted RF': predicted_class_name_rf,\n",
" 'Predicted Centroid': predicted_class_name_centroid,\n",
" 'Correct NN': correct_nn,\n",
" 'Correct RF': correct_rf,\n",
" 'Correct Centroid': correct_centroid\n",
" })\n",
"\n",
"# Convert the list of results into a DataFrame\n",
"results_df = pd.DataFrame(results, columns=columns)\n",
"\n",
"# Print the DataFrame with results\n",
"print(results_df.to_string(index=False))\n",
"\n",
"# Generate a summary\n",
"summary = {\n",
" 'Classifier': ['NN', 'RF', 'Centroid'],\n",
" 'Total Predictions': [len(results_df)] * 3,\n",
" 'Correct Predictions': [\n",
" results_df['Correct NN'].value_counts().get('✅', 0),\n",
" results_df['Correct RF'].value_counts().get('✅', 0),\n",
" results_df['Correct Centroid'].value_counts().get('✅', 0)\n",
" ],\n",
" 'Accuracy': [\n",
" results_df['Correct NN'].value_counts().get('✅', 0) / len(results_df),\n",
" results_df['Correct RF'].value_counts().get('✅', 0) / len(results_df),\n",
" results_df['Correct Centroid'].value_counts().get('✅', 0) / len(results_df)\n",
" ]\n",
"}\n",
"\n",
"summary_df = pd.DataFrame(summary)\n",
"print(\"\\nSummary:\")\n",
"print(summary_df.to_string(index=False))"
],
"id": "eb9f5994c120e928",
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 10ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 10ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 10ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 9ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 8ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
" Text Actual Intent Predicted NN Predicted RF Predicted Centroid Correct NN Correct RF Correct Centroid\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How may I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist you today?\\n\\nCustom Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom Referrals and Specialist Appointments Medical Advice or Symptom Assessment Referrals and Specialist Appointments Medical Advice or Symptom Assessment ❌ ✅ ❌\n",
" Agent 4: Thank you for calling HealthHarbor. This is Agent 4. How can I assist you today?\\n\\nCustomer: Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 4: Good afternoon, thank you for calling HealthHarbor. My name is Agent 4. How can I assist yo Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor. My name is Agent 2. How can I assist you today?\\n\\nCustom Referrals and Specialist Appointments Appointment Scheduling Referrals and Specialist Appointments Referrals and Specialist Appointments ❌ ✅ ✅\n",
" Agent 3: Good morning, thank you for calling HealthHarbor. My name is Agent 3. How can I assist you Follow-up Calls and Care Management Healthcare Facility Information Follow-up Calls and Care Management General Inquiries ❌ ✅ ❌\n",
" Agent 4: Thank you for calling HealthHarbor. My name is Agent 4. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor. My name is Agent 2. How may I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustomer Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How may I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist y Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor. My name is Agent 4. How can I assist you today?\\n\\nCustom Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" Customer: Hello, I would like to speak with someone regarding some medical advice or symptom assessm Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor. My name is Agent 4. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How may I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Referrals and Specialist Appointments Appointment Scheduling Referrals and Specialist Appointments Follow-up Calls and Care Management ❌ ✅ ❌\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom Referrals and Specialist Appointments Medical Advice or Symptom Assessment Referrals and Specialist Appointments Medical Advice or Symptom Assessment ❌ ✅ ❌\n",
" [Phone ringing]\\n\\nAgent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist y Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Referrals and Specialist Appointments Appointment Scheduling Referrals and Specialist Appointments Referrals and Specialist Appointments ❌ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, this is Agent 4 speaking. How may I assist you today?\\n\\n Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How may I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor. My name is Agent 4. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 3: Thank you for calling HealthHarbor. This is Agent 3 speaking. How can I as Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, this is Agent 4 speaking. How can I assist you today?\\n\\n General Inquiries Healthcare Facility Information General Inquiries General Inquiries ❌ ✅ ✅\n",
" Agent: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustomer Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How may I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor. My name is Agent 4. How may I assist you today?\\n\\nCustom Appointment Scheduling Medical Advice or Symptom Assessment Appointment Scheduling Medical Advice or Symptom Assessment ❌ ✅ ❌\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Appointment Scheduling Medical Advice or Symptom Assessment Appointment Scheduling Medical Advice or Symptom Assessment ❌ ✅ ❌\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How may I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor. My name is Agent 1. How can I assist you today?\\n\\nCustom General Inquiries Medical Advice or Symptom Assessment General Inquiries Medical Advice or Symptom Assessment ❌ ✅ ❌\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How may I assist you today?\\n\\nCustom General Inquiries Healthcare Facility Information General Inquiries Healthcare Facility Information ❌ ✅ ❌\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 3: Good morning, thank you for calling HealthHarbor. My name is Agent 3. How can I assist you General Inquiries Healthcare Facility Information General Inquiries Healthcare Facility Information ❌ ✅ ❌\n",
" Agent 4: Thank you for calling HealthHarbor. My name is Agent 4. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist y Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" Call Center Agent: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you toda Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information General Inquiries ✅ ✅ ❌\n",
" [Phone rings]\\n\\nAgent 3: Good morning, thank you for calling HealthHarbor. My name is Agent 3. How ca Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist y Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How may I assist you today?\\n\\nCustom General Inquiries Healthcare Facility Information General Inquiries Healthcare Facility Information ❌ ✅ ❌\n",
" Agent 4: Thank you for calling HealthHarbor. My name is Agent 4. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" [Phone rings]\\n\\nAgent 4: Thank you for calling HealthHarbor, my name is Agent 4. How may I assist you Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information General Inquiries General Inquiries ✅ ❌ ❌\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" Call Center Agent (Agent 1): Thank you for calling HealthHarbor. My name is Agent 1. How may I assis Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 1: Good morning, thank you for calling HealthHarbor. My name is Agent 1. How can I assist you Follow-up Calls and Care Management Lab and Test Results Lab and Test Results Lab and Test Results ❌ ❌ ❌\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How may I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" [Phone rings]\\n\\nAgent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How may I assist you today?\\n\\nCustom Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist y Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How may I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent: Good morning, thank you for calling HealthHarbor. My name is Agent 2. How can I assist you to Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 3: Thank you for calling HealthHarbor, my name is Agent 3. How may I assist y Follow-up Calls and Care Management Insurance and Billing Inquiries Follow-up Calls and Care Management Follow-up Calls and Care Management ❌ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor. My name is Agent 2. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist y Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor. My name is Agent 2. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" Agent 4: Good morning, thank you for calling HealthHarbor. My name is Agent 4. How can I assist you Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom General Inquiries Healthcare Facility Information General Inquiries Healthcare Facility Information ❌ ✅ ❌\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How may I assist you today?\\n\\nCustom Referrals and Specialist Appointments Appointment Scheduling Referrals and Specialist Appointments Appointment Scheduling ❌ ✅ ❌\n",
" Agent 1: Good morning, thank you for calling HealthHarbor. My name is Agent 1. How can I assist you Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor. My name is Agent 1. How can I assist you today?\\n\\nCustom Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 4: Thank you for calling HealthHarbor. My name is Agent 4. How can I assist y General Inquiries Healthcare Facility Information General Inquiries General Inquiries ❌ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist y General Inquiries Medical Advice or Symptom Assessment General Inquiries General Inquiries ❌ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist y Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom General Inquiries Healthcare Facility Information General Inquiries General Inquiries ❌ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor. My name is Agent 1. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
"[Phone ringing]\\n\\nAgent 3: Good morning, this is agent 3 speaking. How may I assist you today?\\n\\nCusto Follow-up Calls and Care Management Medication Refills and Prescription Inquiries Referrals and Specialist Appointments Medication Refills and Prescription Inquiries ❌ ❌ ❌\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How may I assist you today?\\n\\nCustom Follow-up Calls and Care Management Medication Refills and Prescription Inquiries Follow-up Calls and Care Management Referrals and Specialist Appointments ❌ ✅ ❌\n",
" [Phone ringing]\\n\\nAgent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist y Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management Follow-up Calls and Care Management ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist y Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustom Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor. My name is Agent 1. How can I assist you today?\\n\\nCustom General Inquiries Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ❌ ❌ ❌\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom General Inquiries Healthcare Facility Information General Inquiries Healthcare Facility Information ❌ ✅ ❌\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Lab and Test Results Lab and Test Results Lab and Test Results Lab and Test Results ✅ ✅ ✅\n",
" Transcript:\\n\\nAgent 1: Thank you for calling HealthHarbor. My name is Agent 1. How can I assist you t Appointment Scheduling Appointment Scheduling Appointment Scheduling Referrals and Specialist Appointments ✅ ✅ ❌\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 4: Good morning, thank you for calling HealthHarbor. My name is Agent 4. How Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments ✅ ✅ ✅\n",
" Agent 4: Thank you for calling HealthHarbor. My name is Agent 4. How may I assist you today?\\n\\nCustom Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" [Phone rings]\\n\\nAgent 4: Good morning, thank you for calling HealthHarbor. My name is Agent 4. How ma Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries Medication Refills and Prescription Inquiries ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor. My name is Agent 2. How can I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist y Follow-up Calls and Care Management Insurance and Billing Inquiries Follow-up Calls and Care Management Follow-up Calls and Care Management ❌ ✅ ✅\n",
" [Phone ringing]\\n\\nAgent 4: Thank you for calling HealthHarbor, my name is Agent 4. How can I assist y Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 3: Thank you for calling HealthHarbor. My name is Agent 3. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 4: Good morning, thank you for calling HealthHarbor. My name is Agent 4. How can I assist you General Inquiries Insurance and Billing Inquiries General Inquiries Insurance and Billing Inquiries ❌ ✅ ❌\n",
" Agent 2: Thank you for calling HealthHarbor. My name is Agent 2. How can I assist you today?\\n\\nCustom Follow-up Calls and Care Management Lab and Test Results Follow-up Calls and Care Management Follow-up Calls and Care Management ❌ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Healthcare Facility Information Healthcare Facility Information Healthcare Facility Information General Inquiries ✅ ✅ ❌\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How may I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" Agent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you today?\\n\\nCustom Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries Insurance and Billing Inquiries ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent: Thank you for calling HealthHarbor. My name is Agent 2. How can I assist you today?\\n\\nCustomer Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" [Phone rings]\\n\\nAgent 1: Thank you for calling HealthHarbor, my name is Agent 1. How can I assist you Appointment Scheduling Appointment Scheduling Appointment Scheduling Appointment Scheduling ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist you today?\\n\\nCustom Medical Records Requests Medical Records Requests Medical Records Requests Medical Records Requests ✅ ✅ ✅\n",
" Agent 2: Thank you for calling HealthHarbor. My name is Agent 2. How can I assist you today?\\n\\nCustom Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment Medical Advice or Symptom Assessment ✅ ✅ ✅\n",
" [Phone rings]\\n\\nAgent 3: Thank you for calling HealthHarbor, my name is Agent 3. How may I assist you General Inquiries Healthcare Facility Information General Inquiries Healthcare Facility Information ❌ ✅ ❌\n",
" [Phone ringing]\\n\\nAgent 2: Thank you for calling HealthHarbor, my name is Agent 2. How can I assist y Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments Referrals and Specialist Appointments ✅ ✅ ✅\n",
"\n",
"Summary:\n",
"Classifier Total Predictions Correct Predictions Accuracy\n",
" NN 130 102 0.784615\n",
" RF 130 126 0.969231\n",
" Centroid 130 107 0.823077\n"
]
}
],
"execution_count": 47
},
{
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T17:26:24.651137Z",
"start_time": "2024-09-29T17:26:24.648863Z"
}
},
"cell_type": "code",
"source": "test_data[0][\"cleaned_text\"]",
"id": "74fc797696bd02ca",
"outputs": [
{
"data": {
"text/plain": [
"\"Agent 3: Thank you for calling HealthHarbor, my name is Agent 3. How can I assist you today?\\n\\nCustomer: Hi, Agent 3. My name is Emily Johnson. I'm feeling quite worried and I need some information about your healthcare facility.\\n\\nAgent 3: Hello, Emily. I understand that feeling worried can be overwhelming. Don't worry, I'm here to help. Could you let me know what specific information you are looking for?\\n\\nCustomer: Well, I recently moved to this area and I'm looking for a reliable hospital for my family's healthcare needs. I would like to know more about the services HealthHarbor offers.\\n\\nAgent 3: Of course, Emily. HealthHarbor is a full-service hospital that offers a wide range of medical services. We have various departments such as cardiology, pediatrics, orthopedics, and many more. Our goal is to provide comprehensive care to patients in need.\\n\\nCustomer: That sounds great. Can you tell me more about the facilities and equipment available at HealthHarbor?\\n\\nAgent 3: Absolutely. HealthHarbor is equipped with state-of-the-art facilities and the latest medical equipment. We have advanced imaging technology, including MRI and CT scan machines, to ensure accurate diagnosis. Our operating rooms are equipped with modern surgical equipment, and we have a fully-equipped laboratory for diagnostic testing.\\n\\nCustomer: That's reassuring to hear. What about the medical staff at HealthHarbor? Are they qualified and experienced?\\n\\nAgent 3: Our medical staff consists of highly qualified and experienced healthcare professionals. We have a team of skilled doctors, nurses, and specialists who are dedicated to providing the best possible care to our patients. We also ensure that our staff undergoes continuous training to stay updated with the latest medical advancements.\\n\\nCustomer: That's great to know. Another important factor for me is the availability of emergency services. Does HealthHarbor have a 24/7 emergency department?\\n\\nAgent 3: Yes, Emily, we do have a 24/7 emergency department. We understand that emergencies can happen at any time, and our team is always ready to provide immediate medical attention. Our emergency department is staffed with experienced doctors and nurses who are trained to handle all types of emergencies.\\n\\nCustomer: That's a relief. One last thing, do you accept insurance plans? I want to make sure my family's healthcare expenses are covered.\\n\\nAgent 3: Absolutely, Emily. We work with a wide range of insurance providers to ensure that our patients can receive the care they need without any financial burden. I would recommend contacting your insurance provider to check if HealthHarbor is in-network for your specific plan.\\n\\nCustomer: Thank you, Agent 3. You have been very helpful in addressing my concerns. I feel much better now knowing that HealthHarbor is a reliable healthcare facility.\\n\\nAgent 3: You're welcome, Emily. I'm glad I could assist you. If you have any more questions or need further information, feel free to reach out to us anytime. Take care and have a great day!\\n\\nCustomer: Thank you, Agent 3. I appreciate your help. Have a great day too! Goodbye!\\n\\nAgent 3: Goodbye, Emily. Take care!\""
]
},
"execution_count": 57,
"metadata": {},
"output_type": "execute_result"
}
],
"execution_count": 57
},
{
"cell_type": "code",
"id": "e5280026-3540-4f84-a4b1-adc8032f7fee",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:52:10.936599Z",
"start_time": "2024-09-29T15:52:09.474329Z"
}
},
"source": [
"text = test_data[0][\"cleaned_text\"]\n",
"\n",
"start_time = time.time()\n",
"predicted_class_name_nn = predict_class_name_nn(text, embedding_model, nn_classifier)\n",
"prediction_time_nn = time.time() - start_time\n",
"print(f\"Prediction Time: {prediction_time_nn:.4f} seconds\")\n",
"print(f\"Predicted class name: {predicted_class_name_nn}\")"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001B[1m1/1\u001B[0m \u001B[32m━━━━━━━━━━━━━━━━━━━━\u001B[0m\u001B[37m\u001B[0m \u001B[1m0s\u001B[0m 7ms/step\n",
"Prediction Time: 1.4598 seconds\n",
"Predicted class name: Healthcare Facility Information\n"
]
}
],
"execution_count": 48
},
{
"cell_type": "code",
"id": "4baa665c-a273-4a08-9315-a13b8dd63c29",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:55:30.720550Z",
"start_time": "2024-09-29T15:55:28.940846Z"
}
},
"source": [
"start_time = time.time()\n",
"predicted_class_name_rf = predict_class_name_rf(text, embedding_model, rf_classifier)\n",
"prediction_time_rf = time.time() - start_time\n",
"print(f\"Predicted class name: {predicted_class_name_rf}\")\n",
"print(f\"Prediction Time: {prediction_time_rf:.4f} seconds\")"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Predicted class name: Healthcare Facility Information\n",
"Prediction Time: 1.7776 seconds\n"
]
}
],
"execution_count": 51
},
{
"cell_type": "code",
"id": "6a035e2f-4bd2-4748-8863-7b95ecd80592",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:55:33.638589Z",
"start_time": "2024-09-29T15:55:31.630603Z"
}
},
"source": [
"start_time = time.time()\n",
"predicted_class_name_centroid = predict_class_name_centroid(text, centroids, embedding_model)\n",
"prediction_time_centroid = time.time() - start_time\n",
"print(f\"The sentence is classified as: {predicted_class_name_centroid}\")\n",
"print(f\"Prediction Time: {prediction_time_centroid:.4f} seconds\")"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"The sentence is classified as: Healthcare Facility Information\n",
"Prediction Time: 2.0057 seconds\n"
]
}
],
"execution_count": 52
},
{
"cell_type": "code",
"id": "f766d864-140b-4e9f-8403-b48dda92dbdb",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:55:33.720214Z",
"start_time": "2024-09-29T15:55:33.716362Z"
}
},
"source": [
"comparison_df = pd.DataFrame({\n",
" 'Model': ['Neural Network', 'Random Forest', 'Centroid-Based'],\n",
" 'Training Time (s)': [training_time_nn, training_time_rf, centroid_calculation_time],\n",
" 'Prediction Time (s)': [prediction_time_nn, prediction_time_rf, prediction_time_centroid],\n",
" 'Accuracy': [accuracy_nn, accuracy_rf, accuracy_centroid]\n",
"})\n",
"\n",
"print(comparison_df)"
],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" Model Training Time (s) Prediction Time (s) Accuracy\n",
"0 Neural Network 0.883566 1.459820 0.763359\n",
"1 Random Forest 0.773026 1.777647 0.885496\n",
"2 Centroid-Based 0.003611 2.005687 0.854962\n"
]
}
],
"execution_count": 53
},
{
"cell_type": "code",
"id": "ec860f74-c794-46f7-a2c0-235889030839",
"metadata": {
"ExecuteTime": {
"end_time": "2024-09-29T15:57:47.749595Z",
"start_time": "2024-09-29T15:57:45.438265Z"
}
},
"source": [
"# Extract embeddings and labels\n",
"embeddings = np.array(df_train['embeddings'].tolist())\n",
"labels = df_train['intent']\n",
"\n",
"# Perform PCA to reduce dimensions to a more manageable number (e.g., 50)\n",
"pca = PCA(n_components=50, random_state=42)\n",
"embeddings_pca = pca.fit_transform(embeddings)\n",
"\n",
"# Set perplexity to a value less than the number of samples\n",
"perplexity_value = min(30, len(embeddings_pca) - 1)\n",
"\n",
"# Apply t-SNE on the PCA-reduced embeddings\n",
"tsne = TSNE(n_components=2, random_state=42, perplexity=perplexity_value)\n",
"embeddings_2d = tsne.fit_transform(embeddings_pca)\n",
"\n",
"# Create a color map for the classes\n",
"class_names = labels.unique()\n",
"color_map = {class_name: idx for idx, class_name in enumerate(class_names)}\n",
"colors = np.array([color_map[label] for label in labels])\n",
"\n",
"# Plotting\n",
"plt.figure(figsize=(12, 8))\n",
"\n",
"# Plot each class in a different color\n",
"for class_name in class_names:\n",
" indices = np.where(labels == class_name)[0]\n",
" plt.scatter(embeddings_2d[indices, 0], embeddings_2d[indices, 1],\n",
" label=class_name,\n",
" alpha=0.7,\n",
" edgecolors='w')\n",
"\n",
"plt.title('2D t-SNE Visualization after PCA Reduction')\n",
"plt.xlabel('t-SNE Dimension 1')\n",
"plt.ylabel('t-SNE Dimension 2')\n",
"plt.legend()\n",
"plt.grid(True)\n",
"plt.show()"
],
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
"To disable this warning, you can either:\n",
"\t- Avoid using `tokenizers` before the fork if possible\n",
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
]
},
{
"data": {
"text/plain": [
"<Figure size 1200x800 with 1 Axes>"
],
"image/png": "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"
},
"metadata": {},
"output_type": "display_data"
}
],
"execution_count": 54
},
{
"cell_type": "code",
"execution_count": null,
"id": "09aa2e15-c3ec-4db1-9364-5fb88566db3b",
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.10"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment