Neural Ordinary Differential Equation based Recurrent Neural Network Model

TL;DR

This paper introduces two novel ODE-based RNN models, GRU-ODE and LSTM-ODE, which efficiently handle continuous time series with irregular sampling, reducing training time and improving accuracy over existing Neural ODE systems.

Contribution

The paper presents the design of two new ODE-based RNN models that compute states at any time point, simplifying neural network design and reducing computational overhead.

Findings

Require less training time than Latent ODEs and Neural ODEs

Achieve higher accuracy quickly on continuous time series

Simpler neural network design compared to previous Neural ODE systems

Abstract

Neural differential equations are a promising new member in the neural network family. They show the potential of differential equations for time series data analysis. In this paper, the strength of the ordinary differential equation (ODE) is explored with a new extension. The main goal of this work is to answer the following questions: (i)~can ODE be used to redefine the existing neural network model? (ii)~can Neural ODEs solve the irregular sampling rate challenge of existing neural network models for a continuous time series, i.e., length and dynamic nature, (iii)~how to reduce the training and evaluation time of existing Neural ODE systems? This work leverages the mathematical foundation of ODEs to redesign traditional RNNs such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU). The main contribution of this paper is to illustrate the design of two new ODE-based RNN models (GRU-ODE model and LSTM-ODE) which can compute the hidden state and cell state at any point of time using an ODE solver. These models reduce the computation overhead of hidden state and cell state by a vast amount. The performance evaluation of these two new models for learning continuous time series with irregular sampling rate is then demonstrated. Experiments show that these new ODE based RNN models require less training time than Latent ODEs and conventional Neural ODEs. They can achieve higher accuracy quickly, and the design of the neural network is simpler than, previous neural ODE systems.