Identifying Irregularities Usage of Smart Electric Voucher using Support Vector Machines

Abstract

This paper applies machine-learning approach in identifying the irregularities of customer behaviour to the purchase transaction of smart electric vouchers. The Support Vector Machine (SVM) is used as the classification machine learning to identify the irregularities usage into two classes. The performance of the classification system is evaluated using the 10-fold cross-validation technique. Validation results are measured using accuracy, precision, and recall values. Our implementation results showed the use of SVM method gives very good performances in classifying the electrical consumption behaviour. Experimental results with different amounts of data testing indicated that the SVM method has high degree of accuracy, precision, and recall of 99 to 100%.