My lab’s research focuses on explainable and fairness-aware machine learning, systems informatics, and their application to cancer care outcomes analysis, clinical decision support, quality of life, and health equity. The motivation of our research is to translate advanced machine learning and decision under uncertainty techniques into clinical practice in improving the prevention and cure of cancer. Our research work can be categorized into three major aspects, including bias mitigation for fair cancer prevention and treatment, clinical decision support for personalized adaptive therapy, and causal inference for the reasoning of cancer care outcomes and their disparities. Their details can be summarized as follows:

• Bias mitigation for fair cancer prevention and treatment: We investigate sensitive variables (e.g., race, gender, age, rural/urban) in the collected dataset, develop pre-, inter-, and post-processing methods to remove bias among different sub-groups to develop accurate and unbiased cancer prevention and treatment mechanisms. Also, we are interested in mitigating bias and achieving algorithmic fairness by studying intersectional group fairness, individual fairness, and path-specific counterfactual fairness.
• Clinical decision support for personalized adaptive therapy: We develop clinical decision support systems to improve cancer care outcomes based on longitudinal patient-reported outcomes and wearable activity monitors data (e.g., step count, sedentary time, heart rate, and sleep). In particular, we intend to use machine/deep learning to predict treatment outcomes and employ causal reinforcement learning, non-stationary dynamic Bayesian networks, or sequential decision-making to identify the best interventions along the treatment process.  
• Causal inference for the reasoning of cancer care outcomes and their disparities: We develop direct acyclic graphs, partial ancestral graphs, etc. to explore potential pathways/causal-effect relations among cancer patients’ demographic, biological, genomic, imaging characteristics, healthcare interventions, and cancer care outcomes. In addition, we would like to estimate strengths of causal effects of interest, and reason healthcare outcomes and their associated disparities.

My lab is looking for students who would like to pursue their PhD degree from Department of Industrial and Management Systems Engineering or Department of Electrical Engineering in the College of Engineering at the University of South Florida. Candidates with a background in decision under uncertainty, systems modeling, artificial intelligence/machine learning, or casual inference are encouraged to apply. Interested candidates, please send your CV to Yi.Luo@moffitt.org to inquire more details.