Intermittent time series forecasting: local vs global models

TL;DR

This study compares local and global models for forecasting intermittent time series, finding that neural network-based global models, especially D-Linear with negative binomial distribution, outperform local models in accuracy and efficiency.

Contribution

First comprehensive comparison of local and global models, including neural networks with novel distribution heads, on large-scale intermittent time series datasets.

Findings

D-Linear neural network outperforms local models in accuracy.

Transformers are more computationally demanding and less accurate.

Tweedie distribution best estimates high quantiles, negative binomial offers overall best performance.

Abstract

Intermittent time series, characterised by the presence of a significant amount of zeros, constitute a large percentage of inventory items in supply chain. Probabilistic forecasts are needed to plan the inventory levels; the predictive distribution should cover non-negative values, have a mass in zero and a long upper tail. Intermittent time series are commonly forecast using local models, which are trained individually on each time series. In the last years global models, which are trained on a large collection of time series, have become popular for time series forecasting. Global models are often based on neural networks. However, they have not yet been exhaustively tested on intermittent time series. We carry out the first study comparing state-of-the-art local (iETS, TweedieGP) and global models (D-Linear, DeepAR, Transformers) on intermittent time series. For neural networks models we consider three different distribution heads suitable for intermittent time series: negative binomial, hurdle-shifted negative binomial and Tweedie. We use, for the first time, the last two distribution heads with neural networks. We perform experiments on five large datasets comprising more than 40'000 real-world time series. Among neural networks D-Linear provides best accuracy; it also consistently outperforms the local models. Moreover, it has also low computational requirements. Transformers-based architectures are instead much more computationally demanding and less accurate. Among the distribution heads, the Tweedie provides the best estimates of the highest quantiles, while the negative binomial offers overall the best performance.