The author of this document has limited its availability to on-campus or logged-in CSUSB users only.

Off-campus CSUSB users: To download restricted items, please log in to our proxy server with your MyCoyote username and password.

Date of Award


Document Type

Restricted Project: Campus only access

Degree Name

Master of Science in Information Systems and Technology


Information and Decision Sciences

First Reader/Committee Chair

Conrad Shayo


Electricity theft detection in power grids refers to the process of identifying and mitigating instances where electricity consumers illegally tamper with meters or infrastructure to bypass billing mechanisms and obtain electricity without paying for it. This project aims to receive a Gmail notification anytime theft is detected using the CNN-RF model, which provides higher accuracy. The research questions are: (Q1) What measures can be implemented to ensure consumer privacy protection while maintaining the effectiveness of electricity theft detection strategies? (Q2) How can the time to detect energy theft be reduced by using a hybrid machine learning approach for anomaly detection in smart meter data, without sacrificing accuracy? (Q3) What factors, including grid sources, network topology, customer class, and geographic information, can be exploited to monitor abnormalities in energy consumption patterns? The findings for each question are: (Q1) To protect the privacy of customers, a customized email alert is sent out once theft is confirmed. (Q2) By implementing feature selection and data preprocessing, different hybrid models with varying interpretability and accuracy are tested to find a method which detects theft in a short time. (Q3) A Gmail notification including the geographic location and customer class is sent to a particular organization. The conclusions for each question are: (Q1) To protect the privacy of consumers while ensuring efficient detection of electricity, an email notification system could be put in place to inform the organization, once theft is detected. (Q2) By utilizing feature selection and data preparation, advanced machine learning techniques can speed up the identification of energy theft without sacrificing accuracy. (Q3) By using information from power sources, network structure, customer categories and geographic data, irregularities in energy usage patterns can be effectively monitored.