Learning summary features of time series for likelihood free inference

TL;DR

This paper introduces a data-driven method to automatically learn summary features from univariate time series, improving likelihood-free inference performance over traditional hand-crafted features.

Contribution

It proposes a novel automatic feature learning approach for time series in likelihood-free inference, reducing reliance on domain knowledge and manual feature design.

Findings

Learned features can outperform traditional hand-crafted features.

Method is effective on ARMA and Van der Pol oscillator signals.

Data-driven features match or exceed existing LFI methods.

Abstract

There has been an increasing interest from the scientific community in using likelihood-free inference (LFI) to determine which parameters of a given simulator model could best describe a set of experimental data. Despite exciting recent results and a wide range of possible applications, an important bottleneck of LFI when applied to time series data is the necessity of defining a set of summary features, often hand-tailored based on domain knowledge. In this work, we present a data-driven strategy for automatically learning summary features from univariate time series and apply it to signals generated from autoregressive-moving-average (ARMA) models and the Van der Pol Oscillator. Our results indicate that learning summary features from data can compete and even outperform LFI methods based on hand-crafted values such as autocorrelation coefficients even in the linear case.