Udacity: Deep Learning: Reinforcement Learning Framework

[mbhushan]

Value iteration:

Policy Iteration:

Policy improvement:

Reinforcement Learning: Dynamic Programming

Bellman expectation equation:

State Value Functions:

Gridworld Example:

Stochastic Policy:

Deterministic Policy:

Markov Decision Process:

Markov Decision Process:

Action and States:

Goals and Rewards for humanoids:

What are the states:

Reward Hypothesis:

Continuing Task:

Episodic Task:

Goal of the agent: Maximize cumulative expected Reward:

Reward, State, Action: