Jan Kaniewski getnamo

I implemented the DQN model from this paper: https://www.cs.toronto.edu/~vmnih/docs/dqn.pdf.

I used a simple network with two hidden layers and an output layer, instead of the CNN described in the paper due to the relative simplicity of the Cart-Pole environment compared to Atari games.

Note, that I did not yet implement the target network described in the more recent paper here: https://storage.googleapis.com/deepmind-data/assets/papers/DeepMindNature14236Paper.pdf.

The results vary from run to run, sometimes taking 1000 episodes to solve the problem, and at other times taking only 200 episodes.

	def DepthConversion(PointDepth, f):
	H = PointDepth.shape[0]
	W = PointDepth.shape[1]
	i_c = np.float(H) / 2 - 1
	j_c = np.float(W) / 2 - 1
	columns, rows = np.meshgrid(np.linspace(0, W-1, num=W), np.linspace(0, H-1, num=H))
	DistanceFromCenter = ((rows - i_c)2 + (columns - j_c)2)**(0.5)
	PlaneDepth = PointDepth / (1 + (DistanceFromCenter / f)2)(0.5)
	return PlaneDepth

	const id = '__JUNK__';
	// Public key need to be in PKCS8 format
	// ssh-keygen -e -m PKCS8 -f id_rsa.pub > id_rsa.pkcs8
	const publicKey = fs.readFileSync(path.join(__dirname, 'id_rsa.pkcs8'), { encoding : 'utf8' });
	const privateKey = fs.readFileSync(path.join(__dirname, 'id_rsa'), { encoding : 'utf8' });

	// Sign
	const signer = crypto.createSign('RSA-SHA512');
	signer.update(id);
	const signature = signer.sign(privateKey, 'hex');

	<!DOCTYPE html>
	<html>
	<head>
	<script src="https://fb.me/react-15.1.0.js"></script>
	<script src="https://fb.me/react-dom-15.1.0.js"></script>
	<script src="https://npmcdn.com/expect/umd/expect.min.js"></script>
	<script src="https://npmcdn.com/[email protected]/dist/redux.js"
	<meta charset="utf-8">
	<meta name="viewport" content="width=device-width">
	<title>JS Bin</title>

	<!DOCTYPE html>
	<html>
	<head>
	<script src="https://npmcdn.com/expect/umd/expect.min.js"></script>
	<script src="https://npmcdn.com/[email protected]/dist/redux.js"
	<meta charset="utf-8">
	<meta name="viewport" content="width=device-width">
	<title>JS Bin</title>
	></script>
	</head>

	<!DOCTYPE html>
	<html>
	<head>
	<script src="https://npmcdn.com/expect/umd/expect.min.js"></script>
	<link href="http://extjs.cachefly.net/ext-3.1.0/resources/css/ext-all.css" rel="stylesheet" type="text/css" />
	<meta charset="utf-8">
	<meta name="viewport" content="width=device-width">
	<title>JS Bin</title>
	</head>
	<body>


	/*****************
	* bossFight.js *
	*****************
	*
	* NO FARTHER, DR. EVAL!!!!
	* YOU WILL NOT GET OUT OF HERE ALIVE!!!!
	* IT'S TIME YOU SEE MY TRUE FORM!!!!
	* FACE MY ROBOT WRATH!!!!!
	*/

	/********************
	* crispsContest.js *
	********************
	*
	* The Algorithm is almost in our grasp!
	* At long last, we will definitively establish
	* that 3SAT is solvable in polynomial time. It's
	* been a long, strange journey, but it will all be
	* worth it.
	*

	/*
	* robotMaze.js
	*
	* The blue key is inside a labyrinth, and extracting
	* it will not be easy.
	*
	* It's a good thing that you're a AI expert, or
	* we would have to leave empty-handed.
	*/

	/*
	* robotNav.js
	*
	* The green key is located in a slightly more
	* complicated room. You'll need to get the robot
	* past these obstacles.
	*/

	function startLevel(map) {
	// Hint: you can press R or 5 to "rest" and not move the