Research Code

I support open source software, and ask my students to release their code. I am not very good at updating this page, but it contains some pointers to software I’ve been involved in.

INFLUENCE Project

Codebase for “Distributed Influence-Augmented Local Simulators”:  https://github.com/INFLUENCEorg/DIALS

Codebase for “Influence-Augmented Local Simulators”:  https://github.com/INFLUENCEorg/IALS

Codebase for “Online Planning in POMDPs with Self-Improving Simulators”: https://github.com/INFLUENCEorg/POMCP-SIS

Codebase for “Influence-Augmented Online Planning for Complex Environments”: https://github.com/INFLUENCEorg/IAOP .

Codebase for “Decentralized MCTS via Learned Teammate Models”: environment algorithm

Codebase for “Influence-Aware Memory Architectures for Deep Reinforcement Learning in POMDPs”: https://github.com/INFLUENCEorg/influence-aware-memory

Codebase for “Loss Bounds for Approximate Influence-Based Abstraction”: https://github.com/INFLUENCEorg/Approx-IBA-Planning

MADP

The Multiagent Decision Process Toolbox is a toolbox for planning/learning in multiagent systems.

(See our JMLR paper for a quick overview.)

Factored Bayes-adaptive POMDP code

Sammie Katt has released source code for our FBA-POMCP work, which scales up Bayesian RL for POMDPs.

Github/bitbucket

You can check out my github and bitbucket accounts (but I’m afraid I have mostly private repositories.)