Time series analysis is an old art with applications in many disciplines such as finance, medicine, commerce, and weather. Time series are central to all applications in the Financial Services - consumer and commercial - payments, lending, trading, etc. Understanding time series in the real world is challenging due to complex underlying dynamics. We wish to leverage the successes in underlying methodologies and applications in adjacent fields to learn how they can be applied to finance. This workshop aims to gather theoretical and applied researchers interested in analyzing time series and developing new approaches to process sequential information. Special attention will be given to practical as well as novel representations and learning methods in the diverse group of speakers and attendees.
We will discuss a diverse range of topics in financial time series including but not limited to:
Forecasting
Classification
Clustering
Anomaly detection
Changepoint detection
Non-traditional representations (e.g., deep embedding, images, spectrograms etc.)
Augmenting time series representations with other modalities (e.g., videos, satellite imagery etc.)
Synthetic time series generation