Deep Time Warping for Multiple Time Series Alignment

2025-02-22

AI Summary

This work addresses the degradation in classification performance caused by temporal misalignment and scaling in multi-time-series alignment (MTSA). We propose the first end-to-end deep learning framework for joint synchronization and alignment of multiple time series. Unlike conventional pairwise alignment methods or bioinformatics-inspired multiple sequence alignment (MSA) approaches, our method introduces a novel learnable piecewise-linear deformation function, rigorously enforcing boundary conditions, monotonicity, and continuity. We further design a differentiable deep loss function based on dynamic time warping (DTW) and integrate it with deep convolutional networks and differentiable DTW. Evaluated on all 129 datasets in the UCR 2018 benchmark, our approach achieves significant improvements in both classification accuracy and alignment quality, while substantially reducing average runtime. To the best of our knowledge, this is the first method capable of jointly aligning numeric multi-time-series in an end-to-end differentiable manner, thereby filling a critical gap in the literature.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Multiagent Systems: Multiagent Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Time Series Alignment is a critical task in signal processing with numerous real-world applications. In practice, signals often exhibit temporal shifts and scaling, making classification on raw data prone to errors. This paper introduces a novel approach for Multiple Time Series Alignment (MTSA) leveraging Deep Learning techniques. While most existing methods primarily address Multiple Sequence Alignment (MSA) for protein and DNA sequences, there remains a significant gap in alignment methodologies for numerical time series. Additionally, conventional approaches typically focus on pairwise alignment, whereas our proposed method aligns all signals in a multiple manner (all the signals are aligned together at once). This innovation not only enhances alignment efficiency but also significantly improves computational speed. By decomposing into piece-wise linear sections, we introduce varying levels of complexity into the warping function. Additionally, our method ensures the satisfaction of three warping constraints: boundary, monotonicity, and continuity conditions. The utilization of a deep convolutional network allows us to employ a new loss function, addressing some limitations of Dynamic Time Warping (DTW). Experimental results on the UCR Archive 2018, comprising 129 time series datasets, demonstrate that employing our approach to align signals significantly enhances classification accuracy and warping average and also reduces the run time across the majority of these datasets.

Problem

• Aligns multiple time series simultaneously
• Improves classification accuracy significantly
• Reduces computational runtime efficiently

Innovation

• Deep Learning for alignment
• Multiple signals aligned simultaneously
• Convolutional network with new loss function