English

Planning in Dynamic Environments with Conditional Autoregressive Models

Machine Learning 2018-11-27 v1 Artificial Intelligence Robotics Machine Learning

Abstract

We demonstrate the use of conditional autoregressive generative models (van den Oord et al., 2016a) over a discrete latent space (van den Oord et al., 2017b) for forward planning with MCTS. In order to test this method, we introduce a new environment featuring varying difficulty levels, along with moving goals and obstacles. The combination of high-quality frame generation and classical planning approaches nearly matches true environment performance for our task, demonstrating the usefulness of this method for model-based planning in dynamic environments.

Keywords

Cite

@article{arxiv.1811.10097,
  title  = {Planning in Dynamic Environments with Conditional Autoregressive Models},
  author = {Johanna Hansen and Kyle Kastner and Aaron Courville and Gregory Dudek},
  journal= {arXiv preprint arXiv:1811.10097},
  year   = {2018}
}

Comments

6 pages, 1 figure, in Proceedings of the Prediction and Generative Modeling in Reinforcement Learning Workshop at the International Conference on Machine Learning (ICML) in 2018

R2 v1 2026-06-23T05:27:12.138Z