Abnormal Consumption Detection Based on Density in Campus Card

Abstract

Abnormal consumption detection is of great significance to improve the safety of student’s campus card. In order to detect the abnormal consumption on students’ habits and make the model suitable to detect abnormal record for individual. A model based on density is put forward to detect students’ abnormal consumption. Density-Based Spatial Clustering of Applications with Noise (DBSCAN) and Local Outlier Factor (LOF) are integrated to find the abnormal record and abnormal degree of the outlier is marked in the model. Experiments show that abnormal consumption on time, place, cost and frequency can be detected with good accuracy, the severity of the outlier is distinguished correctly. The model is suitable for different students.