Abstract
Abstract Smart meters allow electricity consumption readings at a high time resolution generating time series that can be investigated to extract valuable insights and detect frauds. Using a dataset with recordings from Chinese consumers, we propose an exploratory data analysis and processing to train several classifiers and assess the results. Good results are obtained with ensemble classifiers such as Random Forest (RF), eXtreme Gradient Boosting (XGB) and Multi-Layer Perceptron (MLP) with two layers and a relatively small number of neurons. Real-consumption dataset daily recorded in China consisting of over 42,000 consumers and over 1,000 days is processed with machine learning ML algorithms or classifiers to distinguish between normal and suspicious consumers. In this paper, we will compare a simple feature engineering method that consists in aggregating the data, calculating distances and density function with no feature engineering, proving that the first approach enhances the results and reduces the utility companies’ costs related to on-site inspections. The results are compared with AUC score and ROC curves as the input data is highly skewed.
Highlights
With conventional meters, the suspicious consumers normally consume half of the month while the other half the meter is bypassed so that on average the consumption is constant over time
Various studies were proposed for fraud detection in electricity consumption recorded by conventional or smart meters
Scenario2 – simple feature engineering Regarding the simple feature engineering method, we proposed to perform melting, grouping and pivoting the datasets to calculate three aggregating functions: mean, standard deviation and median
Summary
The suspicious consumers normally consume half of the month while the other half the meter is bypassed so that on average the consumption is constant over time. This way of stealing is no longer met very often with smart meters, as they allow frequent readings that reveal the consumption irregularities over time. The challenge in detecting electricity frauds with smart-metered data consists of continuously stealing of some of the consumed electricity that is difficult to identify. Various studies were proposed for fraud detection in electricity consumption recorded by conventional or smart meters. Emerging IoT and fog are applied for analysing data steams to identify the irregularities in consumption and carry out verifications eliminating the time gaps between the detection and inspection (Siffer, Fouque, Termier, & Largouet, 2017), (Lyu, Jin, Rajasegarar, He, & Palaniswami, 2017)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Proceedings of the International Conference on Business Excellence
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
