Wednesday, April 15, 2026

Time: 16:10-17:00

venue: Mathematics Building Room 527

Speaker:  Prof. Chi-Shian Dai ( Department of Statistics, NCKU.)

Title:Kernel Regression Utilizing External Information as Constraints

Abstract:

In modern scientific research and practice, the widespread availability of large and varied data sources has shifted the analytical focus from single data sets to integrating multiple data sets. Effectively merging these diverse resources requires addressing critical data-quality challenges, including incomplete or partial covariates, summary-level information, and heterogeneity across different data sets. Overcoming these obstacles is vital for achieving reliable and comprehensive data integration.

In this talk, I will introduce a novel methodology that leverages external information as a constraint in kernel regression models. This framework significantly enhances predictive performance by incorporating partial covariate information, accommodating summary-level data, and managing heterogeneous data sets. This approach paves the way for more robust, accurate, and versatile data integration in a wide range of scientific and applied settings.