Non Technical Loses Detection - Experts Labels vs. Inspection Labels in the Learning Stage

Abstract

Non-technical losses detection is a complex task, with high economic impact. The diversity and big number of consumption records, makes it very important to find an efficient automatic method for detection the largest number of frauds with the least amount of experts' hours involved in preprocessing and inspections. This article analyzes the performance of a strategy based on learning from expert labeling: suspect/no-suspect, with one using inspection labels: fraud/no-fraud. Results show that the proposed framework, suitable for imbalance problems, improves performance in terms of the F measure with inspection labels, avoiding hours of experts labeling.

Optimal and Linear F-Measure Classifiers Applied to Non-technical Losses Detection

“…Biscarri et al [5] seek for outliers, Leon et al [23] use Generalized Rule Induction and Di Martino et al [10] combine CS-SVM , One class SVM, C4.5, and OPF classifiers using various features derived from the consumption. In [34] it is compared the feature sets selected when using different classifiers with two different labelling strategies. Different kinds of features are used among this works, for examples, consumption [5,37], contracted power and consumed ratio [15], Wavelet transformation of the monthly consumption [20], amount of inspections made to each client in one period and average power of the area where the customer resides [2], among others.…”