AdaRNN: Adaptive Learning and Forecasting of Time Series

TL;DR

AdaRNN introduces an adaptive framework for time series forecasting that models and mitigates distribution shifts over time, improving accuracy across various real-world applications.

Contribution

The paper proposes a novel adaptive RNN framework with Temporal Distribution Characterization and Matching algorithms to address distribution shifts in time series forecasting.

Findings

Outperforms existing methods with 2.6% higher classification accuracy

Reduces RMSE by 9.0% in experiments

Extensible to Transformer architectures for enhanced performance

Abstract

Time series has wide applications in the real world and is known to be difficult to forecast. Since its statistical properties change over time, its distribution also changes temporally, which will cause severe distribution shift problem to existing methods. However, it remains unexplored to model the time series in the distribution perspective. In this paper, we term this as Temporal Covariate Shift (TCS). This paper proposes Adaptive RNNs (AdaRNN) to tackle the TCS problem by building an adaptive model that generalizes well on the unseen test data. AdaRNN is sequentially composed of two novel algorithms. First, we propose Temporal Distribution Characterization to better characterize the distribution information in the TS. Second, we propose Temporal Distribution Matching to reduce the distribution mismatch in TS to learn the adaptive TS model. AdaRNN is a general framework with flexible distribution distances integrated. Experiments on human activity recognition, air quality prediction, and financial analysis show that AdaRNN outperforms the latest methods by a classification accuracy of 2.6% and significantly reduces the RMSE by 9.0%. We also show that the temporal distribution matching algorithm can be extended in Transformer structure to boost its performance.