Time Series Source Separation with Slow Flows

Abstract

In this paper, we show that slow feature analysis (SFA), a common time series decomposition method, naturally fits into the flow-based models (FBM) framework, a type of invertible neural latent variable models. Building upon recent advances on blind source separation, we show that such a fit makes the time series decomposition identifiable.

INNF+ Workshop, ICML 2020

References in corpus (3)

• NICE: Non-linear Independent Components Estimation
• Recent Advances in Autoencoder-Based Representation Learning
• Disentanglement by Nonlinear ICA with General Incompressible-flow Networks (GIN)