Parametric Augmentation for Time Series Contrastive Learning

arXiv:2402.10434v1 Announce Type: new
Abstract: Modern techniques like contrastive learning have been effectively used in many areas, including computer vision, natural language processing, and graph-structured data. Creating positive examples that assist the model in learning robust and discriminative representations is a crucial stage in contrastive learning approaches. Usually, preset human intuition directs the selection of relevant data augmentations. Due to patterns that are easily recognized by humans, this rule of thumb works well in the vision and language domains. However, …

abstract arxiv augmentation computer computer vision cs.lg data examples graph human intuition language language processing modern natural natural language natural language processing parametric positive processing robust series stage structured data time series type vision