主题：Recent advances of LASSO in econometrics
主讲人：第45届国际奥林匹克数学竞赛金牌获得者 罗晔
主持人：工商管理学院产业经济研究所 刘忠副教授
时间：2015年6月26日(星期五)下午13:00-15:00
地点： 柳林校区 通博楼A座209会议室
主办单位：工商管理学院 科研处
讲座主要内容：
LASSO (Least Absolute Shrinkage and Selection Operator) is a relatively new methodology in statistics and econometrics. Multiple benefits of LASSO include good statistical property, robustness, model selection and substantial computational advantages. In particular, LASSO is very suitable for large dimensional dataset and large scale models. I will briefly discuss LASSO in various environments, including OLS, Quantile Regression, 2SLS, moment selection in GMM and partially identified models, as well as several implications. Due to the advantages of LASSO compared to traditional model selection methods like AIC and BIC, LASSO can be useful in econometrics, finance, big data analytics and bioinfomatics.
