Variability in Classification Outcomes Based on Fuzzy and Non-fuzzy Input Values: A Case Study

This paper presents a case study on the possibility of achieving similar classification outcomes when different types of input datasets were employed in classification tasks. The datasets used in this study were student academic performance datasets collected from the same source but evaluated using fuzzy or non-fuzz values. Six different methods/algorithms were selected to perform the classification tasks. The results obtained from statistical analysis showed that exist variability in classification outcomes induced from datasets collected from different experts, regardless of the types of datasets employed as the input value. The experimental results also showed that exist significant different between classification outcomes produced by methods/algorithms that employed fuzzy input values with the ones employed non-fuzzy input values.