Event Date Details:
refreshments served at 3:15 p.m
- South Hall 5607F
- Department Seminar Series
Title: 2-D logistic fused lasso on large scale data
Abstract: Lasso is a widely used regression analysis method that yields sparse estimator. As sparse estimators leads to an automated model selection procedure, many variations of lasso have been developed to address various problem settings. Fused lasso, which is one of the variations, utilizes the sparse property to obtain locally clustered estimators. In this work, we investigate 2-D fused lasso in logistic regression problem that detects two-dimensional geographical features of estimators given binary response variables. Despite many beneficial features, fused lasso can be computationally costly, especially when the cost function does not have analytic solution. To address this issue, our proposed method facilitates large scale data analysis by utilizing ADMM algorithm. Also, we extend the method to be applicable to flexible data structures including missing observations scenario. Simulations show promising results.