Self-Supervised Learning of Time Series Representation via Diffusion Process and Imputation-Interpolation-Forecasting Mask

Abstract

Time Series Representation Learning (TSRL) focuses on generating informative representations for various Time Series (TS) modeling tasks. Traditional Self-Supervised Learning (SSL) methods in TSRL fall into four main categories: reconstructive, adversarial, contrastive, and predictive, each with a common challenge of sensitivity to noise and intricate data nuances. Recently, diffusion-based methods have shown advanced generative capabilities. However, they primarily target specific application scenarios like imputation and forecasting, leaving a gap in leveraging diffusion models for generic TSRL. Our work, Time Series Diffusion Embedding (TSDE), bridges this gap as the first diffusion-based SSL TSRL approach. TSDE segments TS data into observed and masked parts using an Imputation-Interpolation-Forecasting (IIF) mask. It applies a trainable embedding function, featuring dual-orthogonal Transformer encoders with a crossover mechanism, to the observed part. We train a reverse diffusion process conditioned on the embeddings, designed to predict noise added to the masked part. Extensive experiments demonstrate TSDE's superiority in imputation, interpolation, forecasting, anomaly detection, classification, and clustering. We also conduct an ablation study, present embedding visualizations, and compare inference speed, further substantiating TSDE's efficiency and validity in learning representations of TS data.

Published as a full paper by KDD 2024 Research Track (12 pages as main paper and 11 pages as appendix). Source code available at https://github.com/llcresearch/TSDE

References in corpus (7)

• Natural Scales in Geographical Patterns
• Detecting Spacecraft Anomalies Using LSTMs and Nonparametric Dynamic Thresholding
• SAITS: Self-Attention-based Imputation for Time Series
• Self-Supervised Representation Learning: Introduction, Advances and Challenges
• CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation
• Maximum Likelihood Training of Score-Based Diffusion Models
• Simulation-Informed Revenue Extrapolation with Confidence Estimate for Scaleup Companies Using Scarce Time-Series Data