METODE BOOTSTRAP AGGREGATING REGRESI LOGISTIK BINER UNTUK KETEPATAN KLASIFIKASI KESEJAHTERAAN RUMAH TANGGA DI KOTA PATI

Welfare is one aspect that is quite important to maintain and foster the social and economic stability. Various studies have been conducted regarding the welfare indicates that many factors affect household welfare. Factors affecting household welfare among other gender of household head, age of household heads, the undertaking of the head of household, number of household members, the primary fuel for cooking, buying experience Raskin and the presence or absence of family members who control use cell phone / HP. In this research study on household welfare classification in Pati with the aim to identify factors that influence household welfare in Pati. From the results of studies using Bootstrap aggregating (Bagging) binary logistic regression obtained three predictor variables that significantly influence household welfare in Pati, namely gender heads of household, number of household members, and mastery of mobile phones with a high degree of accuracy at 79, 87%. Results bagging binary logistic regression analysis with bootstrap replication by 50, 60, 70, 80, 90, 100, 150, 200, 626, and 1000 times indicate that there is consistency on each repetition. Keywords : Classification, Binary Logistic Regression, Bootstrap Aggregatin