Abstract:Transformer-based models have traditionally been the primary focus of research for addressing time series forecasting challenges. However, the emergence of recently introduced high-performance linear models has cast doubt upon the effectiveness of transformer architecture in time series forecasting tasks. Throughout, most Transformer variants have represented time series using time point-wise tokenization, which does not provide sufficient semantic information for the attention mechanism. PatchTST expands the receptive field through patch-wise tokenization, mitigating the problem of inadequate information. However, when confronted with multivariate time series forecasting tasks, it does not consider the potential impact of delays and correlation between variates on prediction performance. The recently proposed iTransformer addresses the issue of misalignment between variates by employing series-wise tokenization, yet its embedding method is limited to shallow temporal feature representation. In this work, we propose the Temporal Feature Enhanced Transformer (TFEformer), which deeply integrates patch-wise and series-wise tokenization to enhance the temporal representation of multivariate tokens. Furthermore, we introduce a multi-scale patch fusion mechanism capable of capturing and adaptively integrating temporal features across multiple resolutions. We also enhanced the FFN module to serve as a temporal feature extractor and introduced variate-wise attention to capture the correlations between variables. Extensive experiments on eight real-world datasets have demonstrated that TFEformer outperforms all existing models, achieving state-of-the-art performance. Through experiments, we have also shown that TFEformer improves transformer-based models with superior generalization ability, better utilization of extended lookback windows, and effective suppression of distribution shifts.

Interval Feature Transformation for Time Series Classification Using Perceptually Important Points

A hybrid important points identification for time series: Financial case

Integrating Data-Driven Segmentation, Local Feature Extraction and Fisher Kernel Encoding to Improve Time Series Classification

TFEformer: Temporal Feature Enhanced Transformer for Multivariate Time Series Forecasting

Multilevel Dynamic Time Warping: A Parameter-Light Method for Fast Time Series Classification

PFC: A Novel Perceptual Features-Based Framework for Time Series Classification.

Trend feature extraction method based on important points in time series

Segment-Based Features for Time Series Classification

The Canonical Interval Forest (CIF) Classifier for Time Series Classification

Time Pattern Reconstruction for Classification of Irregularly Sampled Time Series

Interpretable Feature-Temporal Transformer for Short-Term Wind Power Forecasting with Multivariate Time Series

Piecewise Factorization for Time Series Classification.

A Feature Segment Based Time Series Classification Algorithm

Maximum Information Coefficient Feature Selection Method for Interval-Valued Data

QUANT: A Minimalist Interval Method for Time Series Classification

Class Specific Feature Selection for Interval Valued Data Through Interval K-Means Clustering

An Interpretable Time Series Classification Approach Based on Feature Clustering.

Time Series of Intervals Index Based on Clustering

Feature selection for interval-valued data via FRIC-model

COSTI: a New Classifier for Sequences of Temporal Intervals