The Statistics Online Computational Resource (SOCR) designs and disseminates educational materials, web-services, and advanced methods/tools in probability, statistics, and machine learning. This is a research team led by faculties under the UMich-MDP Progam (University of Michigan Multidisciplinary Design Program) that develops enhanced analysis and visualization toolboxes with an emphasis on "Big Data," which are large, heterogeneous, time-varying, multisource and incomplete dataset that are difficult to analyze and interpret in meaningful ways using classical probability, statistical or algorithmic appoaches.
Division: Data Analytics - MIMIC-III
My Role: Interrogated the MIMIC-III database, a large collection of ~43,000 critical care patients from an ICU in Boston, MA, by utilizing R/Rstudio to digest the vital signs, laboratory results, free-text data, and wavefroms available in this unique dataset and predict clinical outcomes via statistical modeling tools; developed advanced AI/ML dara analytics to address specific biomedical, healthcare, neuroimaging-genectics, and other applications.
- SOCR Websitehttps://www.socr.umich.edu/
- MDP Websitehttps://mdp.engin.umich.edu/
- MIMIC-IIIMIMIC-III Clinical Database
- SkillsR/RStudio, Machine Learning, SQL, Quantitative Analytics
- Sourcegithub.com/SOCR/MIMIC-Analytics
- PublicationAppendix 5: Database SQL Queries
- KeywordsSOCR, Big Data, Data Analytics, Numerical Methods