MULTIVARIATE ADAPTIVE REGRESSION SPLINES (MARS) UNTUK KLASIFIKASI STATUS KERJA DI KABUPATEN DEMAK

Abstract

Unemployment is one of the issues relating to economic activities, public relations and also the problems of humanity. Unemployment also occur in Demak and factors suspected as the cause of unemployment in Demak: gender, area of residence, age, status in the household, marriage status and education. Demak BPS records the number of people looking for work (unemployed) as many as 226 228 people, or 29.55% of the working age population. MARS (Multivariate Adaptive Regression Splines) is one of the methods used for classification. MARS is used for high-dimensional data, which is data that has a number of predictor variables for 3 ≤ n ≤ 20 data used in this study is a secondary data from national labor force survey (SAKERNAS) in 2012 To get the best MARS models performed with by combining Maximum Base Function (BF), Minimal Observation (MO), and Maximum Interaction (MI) by trial and error. MARS model is used to classify employment status in Demak are MARS models (BF = 24, MI = 3, MO = 1). Keywords: Unemployment, Classification, MARS