Forecasting Early with Meta Learning

TL;DR

This paper introduces FEML, a meta learning approach that leverages multiple datasets and adversarial sample generation to improve early-stage time series forecasting.

Contribution

It presents a novel meta learning framework with adversarial augmentation for time series forecasting across datasets.

Findings

FEML outperforms single-task learning methods.

Adversarial augmentation improves forecasting accuracy.

Meta learning across datasets enhances early prediction performance.

Abstract

In the early observation period of a time series, there might be only a few historic observations available to learn a model. However, in cases where an existing prior set of datasets is available, Meta learning methods can be applicable. In this paper, we devise a Meta learning method that exploits samples from additional datasets and learns to augment time series through adversarial learning as an auxiliary task for the target dataset. Our model (FEML), is equipped with a shared Convolutional backbone that learns features for varying length inputs from different datasets and has dataset specific heads to forecast for different output lengths. We show that FEML can meta learn across datasets and by additionally learning on adversarial generated samples as auxiliary samples for the target dataset, it can improve the forecasting performance compared to single task learning, and various solutions adapted from Joint learning, Multi-task learning and classic forecasting baselines.