Junting Wang

Ph.D

Statistics

National Institutes of Health

Research Assistant

• Improve estimation accuracy of edge correlation compared to the existing edge functional connectivity in brain network studies through theoretical reasoning and functional MRI analysis, holding potential for mental disease diagnosis through distinctive connectivity patterns between healthy people and patients. • Enhance phenotype prediction accuracy by reducing the dimension of edge correlation from over 100,000,000 to 10,000,000 through multiple hypothesis testing and PCA and incorporate it to node functional connectivity in machine learning models, including linear regression, logistic regression, lasso, ridge, partial least squares, and random forest, using NumPy, Pandas, Sklearn, and SciPy, reducing prediction error by up to 10%. • Presented research results at Joint Statistical Meetings 2023 under the Section on Statistics in Imaging.