Graduate School of Life and Medical Sciences Biomedical Engineering and Informatics / Medical Life Systems

IN-303, IN-307

Our laboratory advances statistical methodology to make sense of complex, large-scale data. We develop interpretable algorithms and multivariate modeling tools, and we apply them to real-world problems—especially in healthcare and life sciences. The themes below highlight our focus on principled theory, practical implementation, and actionable insights for data-driven decision making. 

・Clustering Methods
Designing algorithms to classify objects or variables from large, complex datasets, with an emphasis on interpretability of the clustering results.

 ・Dimension Reduction & Multivariate Analysis
Studying dimensional reduction methods for multi-block data to extract shared factors across datasets on the same subjects, and reduced-rank regression to improve predictive accuracy.

 ・Treatment Effect Estimation / Personalized Medicine
Developing methods based on causal inference to estimate individualized treatment rules, identifying the most effective therapy for each subject using large observational and clinical data.

 ・Applied Data Analysis & Data Science
Applying the above methods to real-world datasets (e.g., healthcare and life sciences) to derive reliable, actionable insights.