Probabilistic structure discovery in time series data

2016/11/21 by Janz, David, Paige, Brooks, Rainforth, Tom +2
#FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML)

paper · doi:10.48550/arxiv.1611.06863

Abstract

Existing methods for structure discovery in time series data construct interpretable, compositional kernels for Gaussian process regression models. While the learned Gaussian process model provides posterior mean and variance estimates, typically the structure is learned via a greedy optimization procedure. This restricts the space of possible solutions and leads to over-confident uncertainty estimates. We introduce a fully Bayesian approach, inferring a full posterior over structures, which more reliably captures the uncertainty of the model.

Related

• Capturing Structure Implicitly from Time-Series having Limited Data
• Time Series Structure Discovery via Probabilistic Program Synthesis
• Structured Bayesian Gaussian process latent variable model
• Bayesian Structure Learning for Stationary Time Series
• Incremental Structure Discovery of Classification via Sequential Monte Carlo