Skip to content

Prototype for multi-agent reinforcement learning framework for ophaned radiation source search using CTDE actor-critic with CNN neural networks

Notifications You must be signed in to change notification settings

bentotten/RAD-TEAM

Repository files navigation

RAD-TEAM

Notice

This repository is undergoing a major refactor. For thesis code, see the Masters thesis branch.

Prerequisites

  • [Optional] Environment manager:

Setup

  1. Initialize environment
    1. [Option 1] with environment manager

      micromamba create --file environment.yml && micromamba activate rad-team

    2. [Option 2] with pip

      pip install -e .

Training

Dont forget to activate any python environments micromamba activate rad-team

Testing

Run all tests with make test

Clean

Refresh with make clean

Docs

Generate docs with make docs. These will generate in the docs/build/html folder.

Development

See DEVELOPMENT.md

Background

I recommend the following resources

  • Reinforcement Learning by Sutton and Barto
  • Google scholar, search for the latest survey papers

Structure

Utils

This directory contains useful tools including

  • normalizers
  • standardizers

About

Prototype for multi-agent reinforcement learning framework for ophaned radiation source search using CTDE actor-critic with CNN neural networks

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published