Recent Advances in Changepoint Modeling

Changepoint models are widely used to detect abrupt changes in the distribution of sequential data. However, handling spatiotemporal dependencies, high dimensionality, multiple samples with varying correlations, and high levels of noise are ongoing challenges in changepoint modeling. This session will include novel methods designed to tackle these challenges. The session will showcase how these methods contribute to earlier cancer detection, better understanding of emerging financial markets in Latin America, improvements in dynamic pricing strategies, and better detection of coordinated air quality declines across large spatial regions.