Deep Learning for time series classification
We live in a temporally dynamic world where data gathered is often associated with a time of acquisition. Making sense of the available time information and the sequential nature of these data is the central task of time series classification, where labels are assigned to time series. Recent advances in this field have been marked by a shift towards deep learning methods due to their state-of-the-art results in computer vision and natural language processing tasks. In this lecture, we will review deep learning architectures proposed for time series classification, which are based on convolutions, recurrence, and attention mechanisms.