MLDM Monday glmnet package and its extensions

Speaker:

I am an attending physician of Chang Gung Memorial Hospital, Keelung. My specialists are gastrointestinal surgery, trauma and surgical critical care. My research interests include applying techniques of statistical learning/maching learning, data mining, and causal inference to biomedical research.

Abstract:

Penalized regression models provide a statistically appealing method to build prediction models from high-dimensional data structure, where its aim is to simultaneously select features and to the model. Since the introduction of the lasso for linear models, the methodology has been extended to generalized linear regression models and Cox model. In addition to the well-known L1-norm (lasso) and L2-norm (ridge) penalty functions, glmnet package uses the elastic net penalty function, which is a linear combination of the L1- and L2-norms. In my presentation, I will introduce the glmnet package and how to conduct data analysis using the package.