Removing Unclassified Elements in Investigating of Financial Wellbeing Attributes Using Rough-Regression Model

In economics research survey, the causal relationship between independent and dependent attributes has been frequently investigated by using regression linear models. However, not easy to achieve the high R-square value between both attributes if there are too many unclassified elements in data sets. This paper presents removing unclassified elements in conventional regression model using rough sets approximation. The proposed model is address to handle the unclassified academic staffs in data set which less contribution for supporting financial wellbeing decision. The result showed that number of unclassified staff has a positive effect to increase coefficient determination (R-square) value in the regression model. In this case study, the financial wellbeing of academic staff is significantly influenced by two different attributes, namely, financial behavior and financial stress. It also may help decision makers or universities management in improving their staff in financial wellness and wellbeing.