Samuel Tan Email and Phone Number

Research Assistant at the University of Michigan School of Public Health, Instructed by Professor Irina Gaynanova• Conduct independent and team-based research• Maintain the R package “iglu,” used for continuous glucose monitor (CGM) data analysis; fix bugs, updatefunctionalities, and publish new versions.• Curate, summarize, and process CGM data from over 1000 diabetes individuals; contributing to a paper inprogress.• Currently developing a model to predict glucose levels in diabetes patients using large language models.

Academy of Mathematics and Systems Science,Conducted research under Professor Qizhai Li, focusing on multidimensional regression and its applications in Biostatistics. Developed a novel test statistic for regression models, demonstrating its effectiveness through rigorous mathematical proofs and extensive simulations. Successfully applied the new test statistic to clinical data, achieving an improvement in testing power by 50% to 100% in various scenarios.

Strengthened collaboration with Committee members to advance the promotion of data science education at the University of Michigan.Created education resources to enhance the understanding of theoretical knowledge in Data Science.

Member of the Data Science Consulting team for the non-profit organization ISAIC.Enhanced the Apricot Case Management System for ISAIC, Analyzed data to support decision-making on student outcomes, Developed and distributed surveys to FISP alumni, collecting valuable feedback, Provided actionable recommendations to improve the FISP program, focusing on entrepreneurship and personal development. Improved data collection for better grant applications and record keeping.

Contributed to the custom architecture team for the "Real vs. Fake" project at the University of Michigan's largest data science club. Utilized deep learning methods within PyTorch to develop an algorithm capable of distinguishing between real and photoshopped photos. Implemented a Convolutional Neural Network (CNN) as the core framework. Achieved a notable test accuracy averaging from 60-70% in identifying photoshopped images.

led an 8-week program, mentoring students in legal, technical, and musical aspects of radio hosting. Collaborated with 17 Executive Committee Members to organize concerts, analyze listener statistics, allocate funds, and oversee station management. Concurrently, hosted the radio show 'Hibikase,' which is a show dedicated to promoting Asian culture and music through engaging content.

Host of the weekly radio show "Hibikase" at KCSB-FM.Curated playlists celebrating East Asian independent music from China, Taiwan, Japan, and South Korea.Promoted East Asian culture in Santa Barbara County, bridging cultural gaps and enhancing community understanding.Hosted bi/trilingual episodes in English, Chinese, and Japanese.Invited diverse guests to discuss cultural insights and explore music, fostering learning and exchange.Reached an estimated 10k-30k listeners across Santa Barbara, Ventura, and parts of Los Angeles Counties.