Sequential Learning of Regression Models by Penalized Estimation

When data arrive in a sequence of two or more datasets, modeling on the most recent dataset should take previous datasets into account. We specifically investigate a strategy for regression modeling when parameter estimates from previous data can be use...