Fuzzy logic-based outlier detection for bio-medical data

Many bio-medical databases such cohort study data suffer from potential errors involved with human factors like mistyping, overlooking some fields. It is crucial to detect such errors at the data entry stage using some techniques like outlier detection. Because such data lie in high-dimensional space and contain many null values, i.e., missing values, most conventional outlier detections are not a good choice for error detection. This paper proposes a fuzzy logic-based outlier detection technique designed to handle data inconsistency at the entry stage for bio-medical databases. The method takes care of the problems in the perspective of the horizontal and vertical consistencies. The method was implemented for a data entry assisting module of a cohort data collection system. In the pilot study, it was observed that the proposed method could detect potential errors at the data entry time.