Pemodelan Tingkat Pengangguran Terbuka di Kabupaten /Kota Provinsi Sumatera Utara Menggunakan Metode Geographichally Weigthed Regression

Abstract

The increase in the open jobless rate in North Sumatra Province reached 6.30 percent. Due to the distinct characteristics of each region in North Sumatra Province, the open unemployment rate problem illustrates the geographical influence of location and the potential for spatial heterogeneity. Geographically Weighted Regression is a method that can help in dealing with spatial problems. By modeling influencing variables, This study intends to learn more about North Sumatra Province's open unemployment rate in 2021. Local partial GWR testing findings demonstrate that the variable of residents who work during the week has significant spatial heterogeneity or is local in nature. However, the factors of the human development index, labor force participation rate, GRDP rate at constant prices and participation rate are purely global in nature. One model that has spatial heterogeneity is Y Kab.Nias = 7,774 − 0,155X1 + 0,245X2 − 1,667X5 − 0,133X7 − 3,26e−5x8 . a fixed gaussian weighting function was used in this investigation for weighting. According to the criteria for the model's goodness, the findings demonstrate that the GWR model outperforms multiple linear regression analysis (OLS), as evidenced by the values for R2, AIC, MSE, and MAPE.