Exploiting Functional Time Series Prediction for PM2.5 Based on Multivariate Variational Mode Decomposition and Anomaly Detection
Journal of Forecasting · 2026-04 · 45 (3) · 1077-1091
DOI: 10.1002/for.70075 ↗At a glance
Journal article in Journal of Forecasting, published online on 2025-12-09. Written by Zhifu Tao and 3 co-authors. 53 references. Funded by: Anhui Office of Philosophy and Social Science.
Abstract (original language)
This paper presents a novel approach to high‐frequency time series forecasting in the context of functional time series, addressing challenges such as data complexity and outliers. The proposed hybrid model integrates outlier detection, multivariate variational mode decomposition (MVMD), and model pooling to enhance forecasting accuracy. Initially, outliers are identified using the isolation forest technique and subsequently replaced with smoothed values via a sliding window moving average. MVMD is then employed to decompose the time series into high‐, mid‐, and low‐frequency components, based on sample entropy. Discrete daily observations are transformed into functional data using Fourier basis functions, and functional principal component analysis (FPCA) is applied for dimensionality reduction, generating principal component scores and functions. Forecasting is carried out through model pooling, which combines statistical, machine learning, and deep learning techniques to predict the principal component scores. The final prediction is obtained by aggregating the forecasts of the predicted scores and their corresponding principal component functions. Empirical results, based on PM2.5 forecasting, demonstrate that the proposed approach significantly outperforms alternative models, offering valuable contributions to air quality monitoring and informed decision‐making.
Funding
- Anhui Office of Philosophy and Social Science
Cite
APA
Tao, Z., Liu, W., Xu, Q., & Wang, P. (2026). Exploiting Functional Time Series Prediction for PM2.5 Based on Multivariate Variational Mode Decomposition and Anomaly Detection. Journal of Forecasting, 45(3), 1077-1091. https://doi.org/10.1002/for.70075
Vancouver
Tao Z, Liu W, Xu Q, Wang P. Exploiting Functional Time Series Prediction for PM2.5 Based on Multivariate Variational Mode Decomposition and Anomaly Detection. Journal of Forecasting. 2026;45(3):1077-1091. doi:10.1002/for.70075
BibTeX
@article{Tao2026,
author = {Tao, Zhifu and Liu, Weiying and Xu, Qin and Wang, Piao},
title = {Exploiting Functional Time Series Prediction for PM2.5 Based on Multivariate Variational Mode Decomposition and Anomaly Detection},
journal = {Journal of Forecasting},
year = {2026},
volume = {45},
number = {3},
pages = {1077--1091},
doi = {10.1002/for.70075}
}
EUVIMED