Statistical Learning for Complex Biomedical Data: From Single-Cell Biomarkers to Seasonal Infectious Diseases

Biomedical data are often complex, high-dimensional, and noisy, making it challenging to identify meaningful patterns and relationships. This talk presents two approaches that use statistical learning to address these challenges. The first introduces an entropy-guided adaptive density-aware kernel (ADAK) for spectral clustering of single-cell mRNA data from hepatocellular carcinoma. By combining information about local density and shared neighborhoods, this method improves clustering stability, supports local interpretability, and helps identify important biomarker genes linked to specific cell subtypes. The second approach develops a periodic mean-reverting stochastic models for seasonal infectious disease data. This model captures both regular seasonal patterns and random variations, offering a flexible and interpretable alternative to traditional time-series models. Together, these methods show how statistical learning techniques can enhance our understanding of complex biomedical data and improve the analysis of both cancer genomics and epidemiological systems.

Short bio: Fahad is an Assistant Professor in the School of Mathematical and Natural Sciences at Arizona State University. He earned his Ph.D. in Statistics from Texas Tech. Before joining ASU, he was a Postdoctoral Fellow at the University of Colorado School of Medicine and a Machine Learning Fellow with the U.S. Food and Drug Administration (FDA). His research focuses on biostatistics, stochastic modeling, and biomedical data science, with an emphasis on developing statistical learning methods for complex biomedical and healthcare data.

Bio
https://sites.google.com/site/gmfahadbinmostafa