<![CDATA[Poverty is one of the most serious problems and must be addressed immediately. One of the steps to overcome poverty is to identify the factors that influence it. One of the statistical techniques used to examine the relationship between predictor variables and the response variable is regression analysis. An important assumption that must be met in regression analysis is the absence of multicollinearity. Multicollinearity refers to a condition where two or more predictor variables are highly correlated, which can reduce the accuracy of the regression model. Therefore, addressing multicollinearity is essential to obtain a reliable and valid model. Inthis study, two methods were employed Ridge regression and Partial Least Square (PLS) with the aim of overcoming the multicollinearity problem. The R2adj value was used as a comparison criterion to evaluate model performance. Both methods were applied to poverty-related data that exhibited signs of multicollinearity. The R2adj value obtained from the ridge regression model was 68.57%, while the PLS model yielded a higher R2adj value of 75.1% . Based on this comparison, it can be concluded that the PLS model produced more optimal results than ridge regression in addressing multicollinearity in the context of modeling factors that influence poverty levels in West Nusa Tenggara Province.]]>
Copyrights © 2025
