Multivariate time series forecasting recently has received extensive attention with its wide application in finance, transportation, environment, and so on. However, few of the currently developed models have considered the impact of noise on prediction. Since multivariate time series contains multiple subsequences with strong nonlinear fluctuations, it is also difficult to obtain satisfactory prediction results. In this paper, aiming at improving prediction performance, we have proposed a novel ensemble threephase model called adaptive noise reducer-stacked auto-encoder-validating-AdaBoost-based long shortterm memory (ANR-SAE-VALSTM). We start with an introduction of a novel ANR for time series noise elimination. The SAEs are then used to extract features from the de-noised multivariate time series. Finally, we feed the de-noised features into the VALSTM to train an ensemble over-fitting prevention predictor. The proposed model is employed on the Beijing PM2.5 dataset and GEFCom2014 Electricity Price dataset. Compared with other popular models, the proposed model has achieved the best prediction performance in all prediction horizons. In addition, a careful ablation study is conducted to demonstrate the efficiency of our model design. INDEX TERMS Multivariate time series forecasting, adaptive noise reducer, stacked auto-encoders, long short-term memory, validating AdaBoost algorithm.
Multivariate Time Series Forecasting (MTSF) has recently emerged its growing importance in many industries. However, how to reduce the influence of the noise components existing in time series on prediction and extract features effectively are still two challenges in MTSF. This paper focuses on those two challenges and proposes a new prediction method based on decomposition-ensemble framework called adaptive sub-series clustering-stacked residual LSTMs-multi-level attention mechanism (ASC-SRLSTMs-MLAttn). The method consists of three stages: decomposition, prediction, and ensemble. In the decomposition stage, the target series is decomposed by Ensemble Empirical Mode Decomposition into multiple sub-series, which will be clustered and reconstructed by the ASC algorithm to reduce the complexity and the time consumption of prediction. In the prediction stage, the sub-series and correlation series will be fed into SRLSTMs-MLAttn for sub-series prediction. The model is based on the encoder-decoder architecture with stacked residual LSTMs as the encoder, which can effectively capture the dependencies among multi variables and the temporal features from multivariate time series. Besides, a multi-level attention mechanism (MLAttn), which makes full use of the encoding information of the encoder, has been introduced to further improve the prediction performance of the model. In the ensemble stage, the predicted values of each sub-series will be summed to obtain the final prediction of the target time series. We also demonstrate the superiority and effectiveness of our proposed method on four public datasets via the conducted comparison experiment and ablation study. INDEX TERMS Multivariate time series forecasting, decomposition-ensemble framework, adaptive sub-series clustering algorithm, stacked residual long short-term memory, multi-level attention mechanism.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.