چاپ صفحه |  صفحه نخست سامانه |  نویسندگان |  اطلاعات تفضیلی |  دانلود مقاله | دانشگاه علوم پزشکی تبریز |
| نویسنده | نفر چندم مقاله |
|---|---|
| فرزانه حمیدی | اول |
| ندا گیلانی | دوم |
| رضا عربی بلاغی | سوم |
| عنوان | متن |
|---|---|
| خلاصه مقاله | There has been a question to see whether machine learning variable selection is superior to the penalized methods such as the LASSO or Elastic Net. In this paper, we address the problem of variable selection in the high dimensional logistic regression. We have proposed several conditions of the sample size, a number of significant variables and different correlation structures to analyse to question. We conducted extensive simulation studies and the results showed that Boruta machine learning variable selection is superior to the LASSO and Elastic Net in the highly correlated structures. On the other hand, in the case of lower correlation, the statistical approaches perform better than Boruta. This paper ended up with two real ovarian cancer datasets. |
| کلمات کلیدی | Variable Selection, Machin learning, Boruta, LASSO, Elastic Net, Simulation. |
| نام فایل | تاریخ درج فایل | اندازه فایل | دانلود |
|---|---|---|---|
| Comparison of the Machine Learning and Classic Penalized Statistical Methods for the Variable Selection.pdf | 1401/07/12 | 205532 | دانلود |
| KETABCHE.pdf | 1401/06/08 | 9527511 | دانلود |
| ghavahi.pdf | 1401/06/08 | 1035286 | دانلود |
