MemDA: Forecasting Urban Time Series with Memory-based Drift Adaptation

TL;DR

MemDA introduces a novel approach for urban time series forecasting that adaptively handles concept drift by encoding periodicity and making real-time adjustments, improving robustness and transferability.

Contribution

The paper presents a new model that encodes data periodicity and uses a meta-dynamic network for on-the-fly drift adaptation, surpassing existing methods.

Findings

Significantly outperforms state-of-the-art methods on real-world datasets.

Reduces sensitivity of prediction models to distribution changes.

Generalizes well across different prediction backbones.

Abstract

Urban time series data forecasting featuring significant contributions to sustainable development is widely studied as an essential task of the smart city. However, with the dramatic and rapid changes in the world environment, the assumption that data obey Independent Identically Distribution is undermined by the subsequent changes in data distribution, known as concept drift, leading to weak replicability and transferability of the model over unseen data. To address the issue, previous approaches typically retrain the model, forcing it to fit the most recent observed data. However, retraining is problematic in that it leads to model lag, consumption of resources, and model re-invalidation, causing the drift problem to be not well solved in realistic scenarios. In this study, we propose a new urban time series prediction model for the concept drift problem, which encodes the drift by considering the periodicity in the data and makes on-the-fly adjustments to the model based on the drift using a meta-dynamic network. Experiments on real-world datasets show that our design significantly outperforms state-of-the-art methods and can be well generalized to existing prediction backbones by reducing their sensitivity to distribution changes.