Course details

Advanced Reinforcement Learning: Implementation

Advanced Reinforcement Learning: Implementation


Overview/Description
Expected Duration
Lesson Objectives
Course Number
Expertise Level



Overview/Description

In this 11-video course, learners can examine the role of reward and discount factors in reinforcement learning, as well as the multi-armed bandit problem and approaches to solving it for machine learning. You will begin by learning how to install the Markov Decision Policy (MDP) toolbox and implement the Discounted Markov Decision Process using the policy iteration algorithm. Next, examine the role of reward and discount factors in reinforcement learning, and the multi-armed bandit problem and solutions. Learn about dynamic programming, policy evaluation, policy iteration, value iteration, and characteristics of Bellman equation. Then learners will explore reinforcement learning agent components and applications; work with reinforcement learning agents using Keras and OpenAI Gym; describe reinforcement learning algorithms and the reinforcement learning taxonomy defined by OpenAI; and implement deep Q-learning with Keras. Finally, observe how to train deep neural networks (DNN) with reinforcement learning for time series forecasting. In the closing exercise, you will recall approaches for resolving the multi-armed bandit problem, list reinforcement learning agent components, and implement deep Q-learning by using Keras and OpenAI Gym.



Expected Duration (hours)
1.6

Lesson Objectives

Advanced Reinforcement Learning: Implementation

  • discover the key concepts covered in this course
  • install the Markov Decision Policy toolbox and implement the Discounted Markov Decision Process using the policy iteration algorithm
  • recognize the role of reward and discount factors in reinforcement learning
  • describe the multi-armed bandit problem and different approaches of solving this problem
  • describe dynamic programming, policy evaluation, policy iteration, value iteration, and characteristics of Bellman equation
  • list reinforcement learning agent components and reinforcement agent applications
  • work with reinforcement learning agents using Keras and OpenAI Gym
  • describe reinforcement learning algorithms and the reinforcement learning taxonomy defined by OpenAI
  • implement deep reinforcement learning or deep Q-learning using Keras and OpenAI Gym
  • recognize how to train deep neural networks using reinforcement learning for time series forecasting
  • recall approaches for resolving the multi-armed bandit problem, list reinforcement learning agent components, and implement deep Q-learning using Keras and OpenAI Gym
  • Course Number:
    it_mlarlndj_02_enus

    Expertise Level
    Intermediate