Conference Papers Year : 2020

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.

Fichier principal
Vignette du fichier
SlowFlows_ICML_workshop.pdf (389.06 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Licence
Loading...

Dates and versions

hal-02959282 , version 1 (06-10-2020)

Licence

Identifiers

  • HAL Id : hal-02959282 , version 1

Cite

Edouard Pineau, Sébastien Razakarivony, Thomas Bonald. Time Series Source Separation with Slow Flows. ICML Workshop on Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models, 2020, (On line ), France. ⟨hal-02959282⟩
175 View
345 Download

Share

  • More