Mixture-of-Linear-Experts for Long-term Time Series Forecasting

arXiv:2312.06786v3 Announce Type: replace
Abstract: Long-term time series forecasting (LTSF) aims to predict future values of a time series given the past values. The current state-of-the-art (SOTA) on this problem is attained in some cases by linear-centric models, which primarily feature a linear mapping layer. However, due to their inherent simplicity, they are not able to adapt their prediction rules to periodic changes in time series patterns. To address this challenge, we propose a Mixture-of-Experts-style augmentation for linear-centric models and …

abstract art arxiv cases cs.ai cs.lg current experts feature forecasting future however layer linear long-term mapping series simplicity sota state time series time series forecasting type values