Pemodelan harga minyak sayuran menggunakan analisis regresi linear berganda.

Abstract

This study focused on application of multiple regression in modeling vegetable oil prices. Five vegetable oil prices, namely CPO, SBO, CNO, PKO, and RSO have been analysed using monthly oil price data from year 2000. We found that multiple linear regression gave the $R^2$ value of 0.887, meaning 88.7\% of variance in CPO price could be explained by RSO, PKO, and CNO. The $t$-test showed that the parameter estimates is significant at one percent level. This study concluded that multicollinearity and autocorrelation were detected inmuliple linear regression and are needed to be considered in further research.