NORMA eResearch @NCI Library

Non-Technical Electricity Loss: Predicting and Defining correlation of Electricity Theft Determinants Using Machine Learning Algorithms

Kawala, Steven (2020) Non-Technical Electricity Loss: Predicting and Defining correlation of Electricity Theft Determinants Using Machine Learning Algorithms. Masters thesis, Dublin, National College of Ireland.

[img]
Preview
PDF (Master of Science)
Download (1MB) | Preview
[img]
Preview
PDF (Configuration manual)
Download (516kB) | Preview

Abstract

Of the issue related with non-specialized irregularities in electricity usages, different strategies have been put in place for effective administration of non-specialized peculiarities in the electricity industry. The effective and best strategy implemented so far to diminish nonspecialized peculiarities and revenue losses is the utilization of Smart advanced utility meters. This strategy makes deceitful exercises increasingly difficult, and it is simple to identify when such deceitful exercise happens. However, this strategy is not extensively utilized in most countries because of the cost associated with the procurement and installation of the smart meters. This research paper looks at how well can Artificial Intelligent Algorithms be used to predict power theft considering the social and economic determinants behind electricity theft. The research proposes the use of seven models on local area power utilization in China, to improve constant precision on the recognition of nontechnical inconsistencies and save revenue loss from utility companies. The models will distinguish and anticipate malevolent power utilization in real-time and chronicled data with anomalous utilization patterns will be associated with electricity theft. All models were evaluated based not only on the accuracy of the model but also sensitivity, specificity, and the AUC results. The analysis of the results did not only look at the exactness deciding the exhibition of the models to judge the performance of the model but also looked at the proportion of right theft forecast. The analysis is successful in predicting theft of electricity and a clear comparison of the models gave a rank to a promising model that supports the researcher.
Keywords: Feature Selection, Electricity theft, Theft determinant, SVM, SOM, kNN, Logistic regression, Decision Tree, Random Forest, Naïve Bayes.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science

Q Science > QA Mathematics > Computer software
T Technology > T Technology (General) > Information Technology > Computer software
Divisions: School of Computing > Master of Science in FinTech
Depositing User: Dan English
Date Deposited: 29 Jan 2021 16:51
Last Modified: 29 Jan 2021 16:51
URI: http://norma.ncirl.ie/id/eprint/4570

Actions (login required)

View Item View Item