Imbalanced time series are universally found in industrial applications. where the number of normal samples is far larger than that of abnormal cases. Traditional machine learning algorithms. such as support vector machine and convolutional neural networks. https://www.learningepistemology.com/product-category/snoring/