Modeling of Quantile Regression to Know the Factors Affecting the High Spread Api Malaria in Indonesia

yahya matdoan (Program Studi Statistika Universitas Pattimura)

JURNAL MATEMATIKA STATISTIKA DAN KOMPUTASI
Vol. 16 No. 3 (2020): JMSK, MAY, 2020

The OLS method estimation is based on a normal distribution, so it is not appropriate to analyze a number of data that are not symmetrical or contain outliers. Therefore, quantile regression was developed which was not affected by outliers. This study compares quantile regression with OLS in the case of factors affecting malaria in Indonesia. The results show that the value of the Quantil Regression model is 0,832 and the MSE value is 0,182. In addition, the OLS model obtained a value of 0,681 and an MSE value of 0,231. So we get the conclusion that the best model is a quantile regression model. Further results were obtained that the main factors causing the spread of malaria in Indonesia were the factor of livable houses, poor population factors and physician factors.