Our research at glance
From High Dimensional Data to Healthcare Provider Profiling
01
Data Collection
UM-KECC has been maintaining a large ESRD database that was
first assembled in 1988. This ESRD database is the backbone
for our research.
02
Data Transformation and integration
We use SAS, SQL and R to alter the structure and format of
the raw data as needed.
03
Statistical Methods
Develop statistical methods and computational algorithms
which include high dimensonal variable selection, survival
analysis, statistical optimization and causal inference.
04
Healthcare Provider Profiling
Develop methods to measure the performance of health care
providers by supplying interested parties with information
on the outcomes of health care.
our projects & software
Healthcare Provider Profiling
Analysis of Readmissions Data Taking Account of Competing Risks.
https://github.com/UM-KevinHe/provideRFEprovideR: Fixed Effects Logistic Model with High-Dimensional Parameters.
https://cran.r-project.org/package=FEprovideROur Projects and Software
Statistical Methods and Computational Algorithms for Big Data Analysis
Block-Wise Steepest Ascent for Large-Scale Survival Analysis with Time-Varying Effects.
https://github.com/UM-KevinHe/TimeVaryingCoxScalable Proximal Methods for Cause-Specific Hazard Modeling with Time-Varying Coefficients.
https://github.com/UM-KevinHe/surtiver