Learning summary features of time series for likelihood free inference
This work provides a data-driven solution for automatically learning summary features for LFI on time series data, which is a significant improvement for researchers and practitioners who previously relied on domain-specific, hand-crafted features.
This paper addresses the bottleneck in Likelihood-Free Inference (LFI) for time series data, which traditionally requires hand-tailored summary features. The authors propose a data-driven strategy to automatically learn these features from univariate time series, demonstrating that their method can compete with and even outperform LFI methods using hand-crafted features like autocorrelation coefficients, even in linear cases.
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.