SutraNets: Sub-series Autoregressive Networks for Long-Sequence, Probabilistic Forecasting

Shane Bergsma, Tim Zeyl, Lei Guo

Keywords: time series, probabilistic forecasting, autoregressive generative models, neural networks

TL;DR: We present SutraNets, a general method for converting probabilistic forecasting of long-sequence time series into multivariate prediction over lower-frequency sub-series, via an autoregressive (across time and sub-series) generative model.

Abstract: We propose SutraNets, a novel method for neural probabilistic forecasting of long-sequence time series. SutraNets use an autoregressive generative model to factorize the likelihood of long sequences into products of conditional probabilities. When generating long sequences, most autoregressive approaches suffer from harmful error accumulation, as well as challenges in modeling long-distance dependencies. SutraNets treat long, univariate prediction as multivariate prediction over lower-frequency sub-series. Autoregression proceeds across time and across sub-series in order to ensure coherent multivariate (and, hence, high-frequency univariate) outputs. Since sub-series can be generated using fewer steps, SutraNets effectively reduce error accumulation and signal path distances. We find SutraNets to significantly improve forecasting accuracy over competitive alternatives on six real-world datasets, including when we vary the number of sub-series and scale up the depth and width of the underlying sequence models.

Submission Number: 1666