Rafael Cosman RafaelCosman

Comparison of modular_rl TRPO on Fully-supervised and Semi-supervised pendulum tasks.

Algorithm-id: alg_zeXD1NReRguwX4NYBg8Tw

I hereby claim:

To claim this, I am signing this object:

	### From https://gist.github.com/RafaelCosman/9efc8181d37cdfdaaca70907bcd97372 ###

	import tensorflow as tf

	import numpy as np
	np.set_printoptions(precision=3)

	sess = tf.InteractiveSession()

	# use the following to import MNIST:

	//Poisson distribution
	//http://en.wikipedia.org/wiki/Poisson_distribution
	function poisson(expected_value){
	var n = 0, //循环计数
	limit = Math.exp(-expected_value), // e -v，其中v是期望值
	x = Math.random(); //生成 0-1之间随机数
	while(x > limit){
	n++;
	x *= Math.random();;
	}

	<!DOCTYPE html>
	<html>
	<head>
	<link href="style.css" rel="stylesheet" />
	<meta charset=utf-8 />
	<title>Cytoscape.js initialisation</title>
	<script src="http://ajax.googleapis.com/ajax/libs/jquery/1/jquery.min.js"></script>
	<script src="http://cytoscape.github.io/cytoscape.js/api/cytoscape.js-latest/cytoscape.min.js"></script>
	<script src="code.js"></script>
	<style id="jsbin-css">

	{
	/*************************************
	* Welcome to Karel in ProcessingJS!
	*
	* Karel is a great way to learn control structures while solving puzzles. Karel is
	* a robot living in a grid world. It only knows a few things about the world around it,
	* and can only follow a few basic commands: move(), turnLeft(), putBeeper(), and pickBeeper().
	* With these simple commands, you'll be able to do so many cool things!
	*
	* Tell Karel what to do by editting its run function. Scroll down to the bottom to get started!