Merge pull request #123 from BAILOOL/master

Updated link to Sutton's book. Changed Lambda to Gamma in FA

DP/README.md

@@ -28,7 +28,7 @@
 
 **Optional:**
 
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 4: Dynamic Programming
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 4: Dynamic Programming
 
 
 ### Exercises

FA/Q-Learning with Value Function Approximation Solution.ipynb

@@ -3,9 +3,7 @@
   {
    "cell_type": "code",
    "execution_count": 1,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "%matplotlib inline\n",
@@ -31,9 +29,7 @@
   {
    "cell_type": "code",
    "execution_count": 2,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stderr",
@@ -50,9 +46,7 @@
   {
    "cell_type": "code",
    "execution_count": 3,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
@@ -74,7 +68,7 @@
     "scaler = sklearn.preprocessing.StandardScaler()\n",
     "scaler.fit(observation_examples)\n",
     "\n",
-    "# Used to converte a state to a featurizes represenation.\n",
+    "# Used to convert a state to a featurizes represenation.\n",
     "# We use RBF kernels with different variances to cover different parts of the space\n",
     "featurizer = sklearn.pipeline.FeatureUnion([\n",
     "        (\"rbf1\", RBFSampler(gamma=5.0, n_components=100)),\n",
@@ -88,9 +82,7 @@
   {
    "cell_type": "code",
    "execution_count": 4,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "class Estimator():\n",
@@ -151,9 +143,7 @@
   {
    "cell_type": "code",
    "execution_count": 5,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "def make_epsilon_greedy_policy(estimator, epsilon, nA):\n",
@@ -182,9 +172,7 @@
   {
    "cell_type": "code",
    "execution_count": 14,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "def q_learning(env, estimator, num_episodes, discount_factor=1.0, epsilon=0.1, epsilon_decay=1.0):\n",
@@ -196,7 +184,7 @@
     "        env: OpenAI environment.\n",
     "        estimator: Action-Value function estimator\n",
     "        num_episodes: Number of episodes to run for.\n",
-    "        discount_factor: Lambda time discount factor.\n",
+    "        discount_factor: Gamma discount factor.\n",
     "        epsilon: Chance the sample a random action. Float betwen 0 and 1.\n",
     "        epsilon_decay: Each episode, epsilon is decayed by this factor\n",
     "    \n",
@@ -283,9 +271,7 @@
   {
    "cell_type": "code",
    "execution_count": 16,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
@@ -305,9 +291,7 @@
   {
    "cell_type": "code",
    "execution_count": 17,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
@@ -384,9 +368,9 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.5.1"
+   "version": "3.5.2"
   }
  },
  "nbformat": 4,
- "nbformat_minor": 0
+ "nbformat_minor": 1
 }

FA/Q-Learning with Value Function Approximation.ipynb

@@ -4,7 +4,7 @@
    "cell_type": "code",
    "execution_count": 1,
    "metadata": {
-    "collapsed": false
+    "collapsed": true
    },
    "outputs": [],
    "source": [
@@ -31,9 +31,7 @@
   {
    "cell_type": "code",
    "execution_count": 2,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stderr",
@@ -50,9 +48,7 @@
   {
    "cell_type": "code",
    "execution_count": 3,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
@@ -89,7 +85,7 @@
    "cell_type": "code",
    "execution_count": 4,
    "metadata": {
-    "collapsed": false
+    "collapsed": true
    },
    "outputs": [],
    "source": [
@@ -149,7 +145,7 @@
    "cell_type": "code",
    "execution_count": 5,
    "metadata": {
-    "collapsed": false
+    "collapsed": true
    },
    "outputs": [],
    "source": [
@@ -180,7 +176,7 @@
    "cell_type": "code",
    "execution_count": 18,
    "metadata": {
-    "collapsed": false
+    "collapsed": true
    },
    "outputs": [],
    "source": [
@@ -193,7 +189,7 @@
     "        env: OpenAI environment.\n",
     "        estimator: Action-Value function estimator\n",
     "        num_episodes: Number of episodes to run for.\n",
-    "        discount_factor: Lambda time discount factor.\n",
+    "        discount_factor: Gamma discount factor.\n",
     "        epsilon: Chance the sample a random action. Float betwen 0 and 1.\n",
     "        epsilon_decay: Each episode, epsilon is decayed by this factor\n",
     "    \n",
@@ -237,9 +233,7 @@
   {
    "cell_type": "code",
    "execution_count": 20,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "name": "stdout",
@@ -259,9 +253,7 @@
   {
    "cell_type": "code",
    "execution_count": 21,
-   "metadata": {
-    "collapsed": false
-   },
+   "metadata": {},
    "outputs": [
     {
      "data": {
@@ -326,9 +318,9 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.4.3"
+   "version": "3.5.2"
   }
  },
  "nbformat": 4,
- "nbformat_minor": 0
+ "nbformat_minor": 1
 }

FA/README.md

@@ -25,8 +25,8 @@
 **Required:**
 
 - David Silver's RL Course Lecture 6 - Value Function Approximation ([video](https://www.youtube.com/watch?v=UoPei5o4fps), [slides](http://www0.cs.ucl.ac.uk/staff/d.silver/web/Teaching_files/FA.pdf))
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 9: On-policy Prediction with Approximation
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 10: On-policy Control with Approximation
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 9: On-policy Prediction with Approximation
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 10: On-policy Control with Approximation
 
 **Optional:**
 
@@ -35,6 +35,8 @@
 
 ### Exercises
 
+- Get familiar with the [Mountain Car Playground](MountainCar%20Playground.ipynb)
+
 - Solve Mountain Car Problem using Q-Learning with Linear Function Approximation
   - [Exercise](Q-Learning%20with%20Value%20Function%20Approximation.ipynb)
   - [Solution](Q-Learning%20with%20Value%20Function%20Approximation%20Solution.ipynb)

Introduction/README.md

@@ -17,7 +17,7 @@
 
 **Required:**
 
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 1: The Reinforcement Learning Problem
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 1: The Reinforcement Learning Problem
 - David Silver's RL Course Lecture 1 - Introduction to Reinforcement Learning ([video](https://www.youtube.com/watch?v=2pWv7GOvuf0), [slides](http://www0.cs.ucl.ac.uk/staff/d.silver/web/Teaching_files/intro_RL.pdf))
 - [OpenAI Gym Tutorial](https://gym.openai.com/docs)
 

MC/README.md

@@ -26,7 +26,7 @@
 
 **Required:**
 
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 5: Monte Carlo Methods
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 5: Monte Carlo Methods
 
 
 **Optional:**

MDP/README.md

@@ -25,7 +25,7 @@
 
 **Required:**
 
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 3: Finite Markov Decision Processes
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 3: Finite Markov Decision Processes
 - David Silver's RL Course Lecture 2 - Markov Decision Processes ([video](https://www.youtube.com/watch?v=lfHX2hHRMVQ), [slides](http://www0.cs.ucl.ac.uk/staff/d.silver/web/Teaching_files/MDP.pdf))
 
 

PolicyGradient/README.md

@@ -36,7 +36,7 @@
 
 **Optional:**
 
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 11: Policy Gradient Methods (Under Construction)
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 11: Policy Gradient Methods (Under Construction)
 - [Deterministic Policy Gradient Algorithms](http://jmlr.org/proceedings/papers/v32/silver14.pdf)
 - [Deterministic Policy Gradient Algorithms (Talk)](http://techtalks.tv/talks/deterministic-policy-gradient-algorithms/61098/)
 - [Continuous control with deep reinforcement learning](https://arxiv.org/abs/1509.02971)

README.md

@@ -2,7 +2,7 @@
 
 This repository provides code, exercises and solutions for popular Reinforcement Learning algorithms. These are meant to serve as a learning tool to complement the theoretical materials from
 
-- [Reinforcement Learning: An Introduction (2nd Edition)](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf)
+- [Reinforcement Learning: An Introduction (2nd Edition)](http://incompleteideas.net/book/bookdraft2017nov5.pdf)
 - [David Silver's Reinforcement Learning Course](http://www0.cs.ucl.ac.uk/staff/d.silver/web/Teaching.html)
 
 Each folder in corresponds to one or more chapters of the above textbook and/or course. In addition to exercises and solution, each folder also contains a list of learning goals, a brief concept summary, and links to the relevant readings.
@@ -50,7 +50,7 @@ All code is written in Python 3 and uses RL environments from [OpenAI Gym](https
 
 Textbooks:
 
-- [Reinforcement Learning: An Introduction (2nd Edition)](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf)
+- [Reinforcement Learning: An Introduction (2nd Edition)](http://incompleteideas.net/book/bookdraft2017nov5.pdf)
 
 Classes:
 

TD/README.md

@@ -28,14 +28,14 @@
 
 **Required:**
 
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 6: Temporal-Difference Learning
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 6: Temporal-Difference Learning
 - David Silver's RL Course Lecture 4 - Model-Free Prediction ([video](https://www.youtube.com/watch?v=PnHCvfgC_ZA), [slides](http://www0.cs.ucl.ac.uk/staff/d.silver/web/Teaching_files/MC-TD.pdf))
 - David Silver's RL Course Lecture 5 - Model-Free Control ([video](https://www.youtube.com/watch?v=0g4j2k_Ggc4), [slides](http://www0.cs.ucl.ac.uk/staff/d.silver/web/Teaching_files/control.pdf))
 
 **Optional:**
 
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 7: Multi-Step Bootstrapping
-- [Reinforcement Learning: An Introduction](http://incompleteideas.net/sutton/book/bookdraft2017june.pdf) - Chapter 12: Eligibility Traces
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 7: Multi-Step Bootstrapping
+- [Reinforcement Learning: An Introduction](http://incompleteideas.net/book/bookdraft2017nov5.pdf) - Chapter 12: Eligibility Traces
 
 
 ### Exercises