Skip to content
This repository has been archived by the owner on Oct 6, 2022. It is now read-only.

Implementation of the Value Iteration Algorithm from "Reinforcement Learning: An Introduction (Second edition)"

Notifications You must be signed in to change notification settings

aaron-seo/Value-Iteration-Algorithm

Repository files navigation

Value Iteration Algorithm for Reinforcement Learning

An implementation of the value iteration algorithm based on "Reinforcement Learning: An Introduction (Second edition)." Here, we iterate the Q table.

We are using the settings similar to the settings for Figure 17.1 on page 646 in the Artificial Intelligence: A Modern Approach (Third Edition) (But not the same!).

“A simple 4 × 3 environment that presents the agent with a sequential decision problem.”

“The "intended" outcome occurs with probability 0.8, but with probability 0.2 the agent moves at right angles to the intended direction. A collision with a wall results in no movement. The two terminal states have reward +1 and -1, respectively, and all other states have a reward of -0.04." (Artificial Intelligence: A Modern Approach (Third Edition), P646)

About

Implementation of the Value Iteration Algorithm from "Reinforcement Learning: An Introduction (Second edition)"

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages