Partially Observable Process Gym

POMDP wrappers for OpenAI Gym

Certainty-Equivalent Expectation Maximization: a scalable algorithm for system identification of partially observed systems

A robot motion planning simulator that can efficiently navigate partially observable environments using deep learning

Scallable partially observable and/or non-Markovian gridworld for planning or reinforcement learning

Multi-agent pursuit in matrix world (pursuitMW)

Application of Reinforcement Learning algorithms (DQN,DRQN,PPO,A2C) to gym's MountainCar-v0

Grids, mountains, and mysterious problems. Solved with Partially-Observable Markov Decision Procesees. Created at Stanford University, by Pablo Rodriguez Bertorello

Recurrent Policies for Handling Partially Observable Environments

Heuristic Search Value Iteration for One-Sided Partially Observable Stochastic Games

Forward A* and Adaptive A* for navigation in partially known environments

Virtual reality (VR) environment for studying human-agent decision making.

An agent being trained to find its way in a stochastic and partially observable maze using SARSA and Q-Learning algorithms.

Implementation of "A Minimax Learning Approach to Off-Policy Evaluation in Confounded Partially Observable Markov Decision Processes" (ICML)

cleaning agent where the environment is partially observable

An implementation of Genetic Programming using Indexed Memory to solve the Partially Observable CartPole problem