Dynamic Gaussian Mixture based Deep Generative Model For Robust Forecasting on Sparse Multivariate Time Series

TL;DR

This paper introduces a dynamic Gaussian mixture-based deep generative model that effectively forecasts sparse multivariate time series by capturing evolving clustering structures, outperforming existing methods especially under high sparsity.

Contribution

The paper proposes a novel dynamic Gaussian mixture distribution within a deep generative framework to model the evolving latent clusters in sparse multivariate time series.

Findings

Outperforms existing methods on real-life datasets

Effectively models dynamic clustering structures

Robust to high sparsity in time series data

Abstract

Forecasting on sparse multivariate time series (MTS) aims to model the predictors of future values of time series given their incomplete past, which is important for many emerging applications. However, most existing methods process MTS's individually, and do not leverage the dynamic distributions underlying the MTS's, leading to sub-optimal results when the sparsity is high. To address this challenge, we propose a novel generative model, which tracks the transition of latent clusters, instead of isolated feature representations, to achieve robust modeling. It is characterized by a newly designed dynamic Gaussian mixture distribution, which captures the dynamics of clustering structures, and is used for emitting timeseries. The generative model is parameterized by neural networks. A structured inference network is also designed for enabling inductive analysis. A gating mechanism is further introduced to dynamically tune the Gaussian mixture distributions. Extensive experimental results on a variety of real-life datasets demonstrate the effectiveness of our method.