A Co-training Approach for Noisy Time Series Learning

Abstract

In this work, we focus on robust time series representation learning. Our assumption is that real-world time series is noisy and complementary information from different views of the same time series plays an important role while analyzing noisy input. Based on this, we create two views for the input time series through two different encoders. We conduct co-training based contrastive learning iteratively to learn the encoders. Our experiments demonstrate that this co-training approach leads to a significant improvement in performance. Especially, by leveraging the complementary information from different views, our proposed TS-CoT method can mitigate the impact of data noise and corruption. Empirical evaluations on four time series benchmarks in unsupervised and semi-supervised settings reveal that TS-CoT outperforms existing methods. Furthermore, the representations learned by TS-CoT can transfer well to downstream tasks through fine-tuning.

Accepted by CIKM2023

References in corpus (9)

• Proceedings of the 29th International Conference on Machine Learning (ICML-12)
• An Iterative Wavelet Threshold for Signal Denoising
• Mixing Up Contrastive Learning: Self-Supervised Representation Learning for Time Series
• CoST: Contrastive Learning of Disentangled Seasonal-Trend Representations for Time Series Forecasting
• Exploring Contrastive Learning in Human Activity Recognition for Healthcare
• AdaRNN: Adaptive Learning and Forecasting of Time Series
• Unsupervised Time-Series Representation Learning with Iterative Bilinear Temporal-Spectral Fusion
• Residual Correction in Real-Time Traffic Forecasting
• Robust Probabilistic Time Series Forecasting