TreeDRNet:A Robust Deep Model for Long Term Time Series Forecasting

TL;DR

TreeDRNet is a novel neural network architecture that enhances long-term time series forecasting by combining robustness, feature selection, and efficiency, outperforming existing models significantly.

Contribution

Introduces TreeDRNet, a multilayer perceptron-based model with a doubly residual structure and tree-based feature selection for robust, efficient long-term forecasting.

Findings

Reduces prediction errors by 20% to 40% compared to state-of-the-art.

Over 10 times more computationally efficient than transformer-based models.

Demonstrates superior robustness and representation power in empirical studies.

Abstract

Various deep learning models, especially some latest Transformer-based approaches, have greatly improved the state-of-art performance for long-term time series forecasting.However, those transformer-based models suffer a severe deterioration performance with prolonged input length, which prohibits them from using extended historical info.Moreover, these methods tend to handle complex examples in long-term forecasting with increased model complexity, which often leads to a significant increase in computation and less robustness in performance(e.g., overfitting). We propose a novel neural network architecture, called TreeDRNet, for more effective long-term forecasting. Inspired by robust regression, we introduce doubly residual link structure to make prediction more robust.Built upon Kolmogorov-Arnold representation theorem, we explicitly introduce feature selection, model ensemble, and a tree structure to further utilize the extended input sequence, which improves the robustness and representation power of TreeDRNet. Unlike previous deep models for sequential forecasting work, TreeDRNet is built entirely on multilayer perceptron and thus enjoys high computational efficiency. Our extensive empirical studies show that TreeDRNet is significantly more effective than state-of-the-art methods, reducing prediction errors by 20% to 40% for multivariate time series. In particular, TreeDRNet is over 10 times more efficient than transformer-based methods. The code will be released soon.