Few-Shot Forecasting of Time-Series with Heterogeneous Channels

TL;DR

This paper introduces a novel approach for few-shot forecasting of multivariate time series with heterogeneous channels, leveraging meta-learning and deep set models to generalize across diverse datasets.

Contribution

It formalizes the problem of few-shot time-series forecasting with heterogeneous channels and develops a permutation-invariant deep set model incorporating temporal embeddings.

Findings

The proposed model outperforms baseline methods in generalization across datasets.

Assembled the first meta-dataset of 40 multivariate time-series datasets.

Demonstrated effective learning across tasks with heterogeneous channels.

Abstract

Learning complex time series forecasting models usually requires a large amount of data, as each model is trained from scratch for each task/data set. Leveraging learning experience with similar datasets is a well-established technique for classification problems called few-shot classification. However, existing approaches cannot be applied to time-series forecasting because i) multivariate time-series datasets have different channels and ii) forecasting is principally different from classification. In this paper we formalize the problem of few-shot forecasting of time-series with heterogeneous channels for the first time. Extending recent work on heterogeneous attributes in vector data, we develop a model composed of permutation-invariant deep set-blocks which incorporate a temporal embedding. We assemble the first meta-dataset of 40 multivariate time-series datasets and show through experiments that our model provides a good generalization, outperforming baselines carried over from simpler scenarios that either fail to learn across tasks or miss temporal information.