Compactness rate as a rule selection index based on Rough Set Theory to improve data analysis for personal investment portfolios

This study proposes a selection index technique, namely a compactness rate based on Rough Set Theory (RST), for improving data analysis, eliminating data amount and reducing the number of decision rule. This study uses an empirical real-case involving a personal investment portfolio to demonstrate the proposed method. The presented case includes 75 rules generated by the RST. The rules are vague and fragmentary, making it very difficult to interpret the information. Many rules have the same strength and number of support objects and condition parts. These are creating a critical problem for decision making. The new method proposed in this study not only enables the selection of interesting rules, but it also reduces the data amount, and offers alternative strategies that can help decision-makers analyze data.