This course explores the unique characteristics of financial time series, their stylized facts, the significance of
missing data, and alternative data sources, with applications to risk modeling. Students learn parametric
approaches (LSTM, spectral methods, structural causal models) and non-parametric techniques (recurrent
integrations, transformers, generative models) for forecasting returns and risk. The focus is on leveraging all
available data using advanced ML architectures to capture temporal dependencies and market dynamics.
