Evolutionary Planning in Latent Space

Thor Valentin Aakjær Nielsen Olesen, Dennis Thinh Tan Nguyen, Rasmus Berg Palm, Sebastian Risi

Research output: Conference Article in Proceeding or Book/Report chapterArticle in proceedingsResearchpeer-review


Planning is a powerful approach to reinforcement learning with several desirable properties. However, it requires a model of the world, which is not readily available in many real-life problems. In this paper, we propose to learn a world model that enables Evolutionary Planning in Latent Space (EPLS). We use a Variational Auto Encoder (VAE) to learn a compressed latent representation of individual observations and extend a Mixture Density Recurrent Neural Network (MDRNN) to learn a stochastic, multi-modal forward model of the world that can be used for planning. We use the Random Mutation Hill Climbing (RMHC) to find a sequence of actions that maximize expected reward in this learned model of the world. We demonstrate how to build a model of the world by bootstrapping it with rollouts from a random policy and iteratively refining it with rollouts from an increasingly accurate planning policy using the learned world model. After a few iterations of this refinement, our planning agents are better than standard model-free reinforcement learning approaches demonstrating the viability of our approach.
Original languageEnglish
Title of host publicationInternational Conference on the Applications of Evolutionary Computation
Publication date2021
Publication statusPublished - 2021
EventInternational Conference on the Applications of Evolutionary Computation -
Duration: 7 Apr 2021 → …


ConferenceInternational Conference on the Applications of Evolutionary Computation
Period07/04/2021 → …


Dive into the research topics of 'Evolutionary Planning in Latent Space'. Together they form a unique fingerprint.

Cite this