Non-Deterministic Outlier Detection Method Based on the Variable Precision Rough Set Model
Alberto Fernández Oliva, Francisco Maciá Pérez, Jos�Vicente Bern�Martinez, Miguel Alfonso Abreu Ortega · Computer Systems Science and Engineering · 2019
This study presents a method for the detection of outliers based on the Variable Precision Rough Set Model (VPRSM). The basis of this model is the generalisation of the standard concept of a set inclusion relation on which the Rough Set Basic Model (RSBM) is based. The primary contribution of this study is the improvement in detection quality, which is achieved due to the generalisation allowed by the classification system that allows a certain degree of uncertainty. From this method, a computationally efficient algorithm is proposed. The experiments performed with a real scenario and a comparison of the results with the RSBM-based method demonstrate the effectiveness of the method as well as the algorithm’s efficiency in diverse contexts, which also involve large amounts of data.