Learnt dynamics generalizes across tasks, datasets, and populations
This addresses the challenge of scarce training data in dynamic signal classification for applications like neuroimaging and keyword detection, offering a method that generalizes across domains, though it is incremental as it builds on self-supervised techniques.
The paper tackles the problem of classifying multivariate dynamic signals with limited training data by leveraging signal dynamics through self-supervised pre-training, resulting in improved classification performance across various tasks and datasets, such as keyword detection and functional neuroimaging, where a single embedding generalized across disorders, age groups, and datasets.
Differentiating multivariate dynamic signals is a difficult learning problem as the feature space may be large yet often only a few training examples are available. Traditional approaches to this problem either proceed from handcrafted features or require large datasets to combat the m >> n problem. In this paper, we show that the source of the problem---signal dynamics---can be used to our advantage and noticeably improve classification performance on a range of discrimination tasks when training data is scarce. We demonstrate that self-supervised pre-training guided by signal dynamics produces embedding that generalizes across tasks, datasets, data collection sites, and data distributions. We perform an extensive evaluation of this approach on a range of tasks including simulated data, keyword detection problem, and a range of functional neuroimaging data, where we show that a single embedding learnt on healthy subjects generalizes across a number of disorders, age groups, and datasets.