Adaptive Multi-Scale Decomposition Framework for Time Series Forecasting

TL;DR

This paper introduces an MLP-based adaptive multi-scale decomposition framework for time series forecasting that effectively captures complex temporal patterns and dependencies, achieving state-of-the-art results with improved efficiency.

Contribution

The paper proposes a novel MLP-based framework with multi-scale decomposition and specialized blocks to enhance temporal pattern modeling in TSF, addressing limitations of existing methods.

Findings

Achieves state-of-the-art performance in long-term and short-term forecasting.

Demonstrates superior efficiency over Transformer-based methods.

Effectively models complex temporal and channel dependencies.

Abstract

Transformer-based and MLP-based methods have emerged as leading approaches in time series forecasting (TSF). While Transformer-based methods excel in capturing long-range dependencies, they suffer from high computational complexities and tend to overfit. Conversely, MLP-based methods offer computational efficiency and adeptness in modeling temporal dynamics, but they struggle with capturing complex temporal patterns effectively. To address these challenges, we propose a novel MLP-based Adaptive Multi-Scale Decomposition (AMD) framework for TSF. Our framework decomposes time series into distinct temporal patterns at multiple scales, leveraging the Multi-Scale Decomposable Mixing (MDM) block to dissect and aggregate these patterns in a residual manner. Complemented by the Dual Dependency Interaction (DDI) block and the Adaptive Multi-predictor Synthesis (AMS) block, our approach effectively models both temporal and channel dependencies and utilizes autocorrelation to refine multi-scale data integration. Comprehensive experiments demonstrate that our AMD framework not only overcomes the limitations of existing methods but also consistently achieves state-of-the-art performance in both long-term and short-term forecasting tasks across various datasets, showcasing superior efficiency. Code is available at https://github.com/TROUBADOUR000/AMD