Analyzing Irregular Time Series Data with Stable Neural Stochastic Differential Equations

Proposing stable classes of Neural SDEs for analyzing irregular time series data, demonstrating robustness under distribution shift.