22 February ~ Ian Wang (Stats)

Title: Extensions of Lasso

 

Abstract: When Lasso is used as a kind of method for variable selection, there are some limitations or drawbacks engaged with it such as multicollinearity of covariates and overestimation. Thus some optimizations are of interest to the researchers to mitigate these limitations. In this seminar, I will discuss some extensional methods after Lasso to help yield a better regression outcome and some graphs from applications are demonstrated to illustrate the benefit from using them.