HRDF Course Deep Reinforcement Learning for Beginners

Reinforcement Learning is a type of machine learning that allows machines and software agents to act smart and automatically detect the ideal behavior within a specific environment, in order to maximize its performance and productivity. Reinforcement Learning is becoming popular because it not only serves as an way to study how machine and software agents learn to act, it is also been used as a tool for constructing autonomous systems that improve themselves with experience.

This course will teach you the basics of reinforcement learning and implement using Python and Tensorflow. The topics include:

• Q Learning
• Sarsa
• Deep Q-Network
• Open AI Gym
• Policy Gradient
• Actor Critic

Module 1 Introduction to RL

• What is Reinforcement Learning (RL)
• Basic Concepts of RL
• Applications of RL
• RL Approaches
• Key RL Algorithms

Module 2 Q Learning

• What is Q-Learning?
• Q-Value, Discount Factor and Learning Rate
• Q Table Update
• Policy
• Q-Learning Algorithm
• Q-Learning Demo

Module 3 Sarsa

• What is Sarsa?
• Q Value Update
• Sarsa Algorithm
• Sarsa Demo

Module 4 Open AI Gym

• OpenAI Gym
• Install Gym
• Render Gym Env
• Action on Gym Env
• Q-Learning on Gym Env

Module 5 Deep Q-Network

• What is Deep Q Network (DQN)?
• DQN Loss Function
• Experience Replay
• DQN Algorithm
• DQN Demo
• Atari DQN

Module 6 Policy Gradient

• What is Policy Gradient (PG)?
• PG Algorithm
• PG Demo

Module 7 Actor-Critic

• What is Actor-Critic?
• Actor Critic Algorithm
• Actor Critic Demo
• AlphaGo

Dr. Aanand is a Full Stack Data Scientist who once had a torrid love affair with Physics. He has consulted and published in the area of Public Health, Electricity Markets, Telecom, BFSI, Advertising & Communication Strategies and Digital & Social Media Technologies. He has worked on assignments with international agencies such as International Monetary Fund, World Bank, Royal Netherland Embassy etc. besides MNCs like Tata Consultancy Services, Kie Square Consulting and several government organizations of national importance.

He regularly conducts general training programs in Python (Pandas, NumPy, SciPy, Matplotlib, Bokeh), R (dplyr, rstanarm, knitR, ggplot2), Data Visualization (Tableau, D3.js) and Machine Learnng (Reinforced Learning, Scikit Learn) and specialized training programs on Structural Equation Modeling and SAP Hana.

He holds a doctorate in Operations Research from Indian Institute of Management Ahmedabad and a post graduate in Physics from University of Mumbai. He has advanced training in mathematical programming including optimization, advanced multivariate data analysis, and simulation techniques. When he is not teaching or consulting he can be found meditating or heading for an adventurous trek.