Electricity Theft Detection Using Machine Learning Algorithm
Student: Abdullahi Umar (Project, 2025)
Department of Computer Science and Informatics
Kaduna State University, Kaduna, Kaduna State
Abstract
Electricity theft is a widespread problem that plagues both developed and developing countries. It has inflicted substantial revenue losses on the electricity industry, marked by Technical Losses (TL) and Non-Technical Losses (NTL) both in Transmission and Distribution. TL resulted from power dissipation in infrastructure, while prevalent NTL challenges in Nigeria included illegal connections and meter tampering. This research successfully addressed NTL challenges by predicting electricity theft among metered customers, employing the machine learning algorithm for efficient data balancing and classification. The system featured a user-friendly interface designed for ease of use. It encompassed data loading, pre-processing, visualization, and model training and evaluation. The trained model effectively predicted potential theft, with subsequent database interactions concluding the process. This outstanding performance not only positionsthe proposed system as the best choice but also establishes a foundation for a more sustainable future in the electricity industry.
Keywords
For the full publication, please contact the author directly at: umarabdullahi1100@gmail.com
Filters
Institutions
- Mohammed Lawan College of Agriculture, Maiduguri, Borno State 12
- Moshood Abiola Polytechnic, Abeokuta, Ogun State 7
- Nasarawa State University, Keffi, Nasarawa State 8
- Niger Delta University, Wilberforce Island, Bayelsa State 28
- Niger State College of Education, Minna, (Affl To Usmanu Danfodiyo Uni, Sokoto) 1
- Nigeria Maritime University, Okerenkoko, Delta State 1
- Nigerian Army University, Biu, Borno State 3
- Nile University of Nigeria, Abuja 3
- Nnamdi Azikiwe University, Awka, Anambra State 98
- Northwest University, Kano, Kano State 179