Disentangling Structured Components: Towards Adaptive, Interpretable and Scalable Time Series Forecasting

22 May 2023  ยท  Jinliang Deng, Xiusi Chen, Renhe Jiang, Du Yin, Yi Yang, Xuan Song, Ivor W. Tsang ยท

Multivariate time-series (MTS) forecasting is a paramount and fundamental problem in many real-world applications. The core issue in MTS forecasting is how to effectively model complex spatial-temporal patterns. In this paper, we develop a adaptive, interpretable and scalable forecasting framework, which seeks to individually model each component of the spatial-temporal patterns. We name this framework SCNN, as an acronym of Structured Component-based Neural Network. SCNN works with a pre-defined generative process of MTS, which arithmetically characterizes the latent structure of the spatial-temporal patterns. In line with its reverse process, SCNN decouples MTS data into structured and heterogeneous components and then respectively extrapolates the evolution of these components, the dynamics of which are more traceable and predictable than the original MTS. Extensive experiments are conducted to demonstrate that SCNN can achieve superior performance over state-of-the-art models on three real-world datasets. Additionally, we examine SCNN with different configurations and perform in-depth analyses of the properties of SCNN.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Time Series Forecasting ETTh1 (192) Multivariate SCNN MSE 0.379 # 2
MAE 0.398 # 11
Time Series Forecasting ETTm1 (192) Multivariate SCNN MSE 0.327 # 3
Time Series Forecasting ETTm1 (96) Multivariate SCNN MSE 0.287 # 4
Time Series Forecasting ETTm2 (192) Multivariate SCNN MSE 0.221 # 5
Time Series Forecasting ETTm2 (96) Multivariate SCNN MSE 0.163 # 4
Time Series Forecasting Weather (192) SCNN MSE 0.188 # 3
Time Series Forecasting Weather (96) SCNN MSE 0.142 # 1

Methods