Added project on Reinforcement learning using OpenAI gym

reinforcement_learning/data/readme.md

@@ -0,0 +1,3 @@
+The data being used is stored in the following format :-
+"state,action,reward,next_state,done" based on the Q-learning strategy after learning from experiences played in the episodes.
+The output weights are stored in the solution folder named "cartpole_model".

reinforcement_learning/exercise/readme.md

@@ -0,0 +1,14 @@
+# Exercise goal
+- Learn the concept of Reinforcement learning by solving the cartpole player problem with the use of OpenAI gym.
+
+# Data
+- Using the cartpole player from OpenAI gym. [Link](https://gym.openai.com/docs/)
+
+# Task
+- This exercise helps in learning reinforcement learning using 3 approaches. 
+- First, using random playing strategy multiple times and saving parameters for the best average episode. 
+- Second, playing multiple times and updating the weights which leads to best optimum time.
+- Third, using Deep Q-Learning strategy to optimize the parameters, left and right actions per se for various frames.
+
+# Prerequisites
+- Must have experience with basic Deep Learning using Tensorflow.