I. Course Goal：

This course covers some of the statistical foundations for medical AI. In particular, the topics include linear regression, logistic regression, ROC curve analysis, Poisson regression, and survival analysis. The course will start from continuous data to illustrate the rationale of the linear regression model and its applications. The concept will be extended to different data structures, including binary data, count data, and survival data. Another topic of this course will focus on the implementation of the R software and the ability to analyze the real data sets. We expect students can choose the appropriate methods (as well as understand the rationale of the methods) to conduct the statistical analysis for a given data set correctly.

II. Contents：

• Review basic statistical analysis
• Simple linear regression
• Least squares estimate
• R-square & adjusted R-square
• mean response, prediction, residual
• Multiple linear regression
• Full & Reduced model, ANOVA
• Introduction of generalized linear model, MLE
• Logistic regression
• ROC curve analysis
• Poisson regression
• Introduction of survival analysis
• Kaplan-Meier estimator
• Log rank test
• Cox PH model and its extensions