0

Disentangled Generative Models for Robust Prediction of System Dynamics

Using supervised disentanglement in generative models improves handling of out-of-distribution generalization and long-term stability in dynamical system predictions.

Year
2021
Venue
disentangled-generative-models-for-robust
Authors
6
Hosting
Abstract onlyARXIV-DEFAULT

Cite

Notes

Only stored in your browser.

Attribution

Abstract & full text
arxiv.org/abs/2108.11684v3ARXIV-DEFAULT
TL;DR
Semantic Scholar
Attribution policy →

Abstract

Deep neural networks have become increasingly of interest in dynamical system prediction, but out-of-distribution generalization and long-term stability still remains challenging. In this work, we treat the domain parameters of dynamical systems as factors of variation of the data generating process. By leveraging ideas from supervised disentanglement and causal factorization, we aim to separate the domain parameters from the dynamics in the latent space of generative models. In our experiments we model dynamics both in phase space and in video sequences and conduct rigorous OOD evaluations. Results indicate that disentangled VAEs adapt better to domain parameters spaces that were not present in the training data. At the same time, disentanglement can improve the long-term and out-of-distribution predictions of state-of-the-art models in video sequences.

Authors

6