Inference, Prediction, and Entropy-Rate Estimation of Continuous-time, Discrete-event Processes

TL;DR

This paper introduces new methods for inferring, predicting, and estimating the entropy rate of continuous-time, discrete-event processes using neural network-based Bayesian structural inference, demonstrating competitive performance on synthetic data.

Contribution

It extends Bayesian structural inference to continuous-time processes using neural networks, a novel approach for this class of problems.

Findings

Methods are competitive with state-of-the-art for prediction.

Methods effectively estimate entropy rates.

Applicable to complex synthetic data.

Abstract

Inferring models, predicting the future, and estimating the entropy rate of discrete-time, discrete-event processes is well-worn ground. However, a much broader class of discrete-event processes operates in continuous-time. Here, we provide new methods for inferring, predicting, and estimating them. The methods rely on an extension of Bayesian structural inference that takes advantage of neural network's universal approximation power. Based on experiments with complex synthetic data, the methods are competitive with the state-of-the-art for prediction and entropy-rate estimation.