Piecewise-constant Neural ODEs
arXiv:2106.06621
Abstract
Neural networks are a popular tool for modeling sequential data but they generally do not treat time as a continuous variable. Neural ODEs represent an important exception: they parameterize the time derivative of a hidden state with a neural network and then integrate over arbitrary amounts of time. But these parameterizations, which have arbitrary curvature, can be hard to integrate and thus train and evaluate. In this paper, we propose making a piecewise-constant approximation to Neural ODEs to mitigate these issues. Our model can be integrated exactly via Euler integration and can generate autoregressive samples in 3-20 times fewer steps than comparable RNN and ODE-RNN models. We evaluate our model on several synthetic physics tasks and a planning task inspired by the game of billiards. We find that it matches the performance of baseline approaches while requiring less time to train and evaluate.
arXiv comment8 pages, 5 figures (not counting appendix)
References in corpus (7)
- Character-based Neural Machine Translation
- Latent ODEs for Irregularly-Sampled Time Series
- Neural Jump Stochastic Differential Equations
- How to train your neural ODE: the world of Jacobian and kinetic regularization
- Recurrent Independent Mechanisms
- Discrete Event, Continuous Time RNNs
- Learning Neural Event Functions for Ordinary Differential Equations