Premarital Sex Behavior Model with Generalized Linear Mixed Model Least Absolute Shrinkage and Selection Operator dan Generalized Linear Mixed Model GROUP Least Absolute Shrinkage and Selection Operator

ABSTRACT
 Premarital sexual behavior is that carried out between men and women without legal marriage. As the number of premarital sex increases, efforts need to take. One can do identify main factors contributing reducing or increasing by a Regression model. In context behavior, environmental influences cannot be ignored. GLMM used model data grouped into certain Groups, include environment effect modeled as mixed in GLMM. terms parsimony, LASSO method selection variables. This research uses Group approach data. The best describes South Sulawesi based on greatest AUC value. variables significantly influence are Type Residence (X_1), Education Level (X_2), Literacy (X_3), Internet use (X_4), Knowledge Contraceptive Methods (X_6), Health Insurance Ownership (X_7), Employment Status (X_8), Sexually Transmitted Diseases (X_9). By knowing government expected take appropriate action for handling it.