Assessing Reported Electrical Accidents Using Regression and Artificial Neural Network Methods


  • Beena Puthillath, M. Bhasi, C A Babu


Electrical Accidents, Artificial Neural Network, Multilayer Perception, Radial Bias Function, Accident Model.


Most of the electrical safety issues are accidental since people’s exposure to the source of electricity and electrical appliances are increasing. Injury or death due to electrical contact is common due to the extensive use of electricity in various sectors. So there is a need to investigate electrocution accidents. The investigation is not only to understand or detect the cause of accidents but also to prevent electrocution issues in near future. The author aims to understand and model the reported electrical accidents. In this study, the author tried to put various causes of reported electrical accidents in the framework of the Regression Model and Artificial Neural Network Model. Data were collected from Indian statistics for understanding the cause of accidents and to develop the Regression and ANN Models. In the regression model only known factors are considered in modelling electrical accidents. In ANN Model both known and unknown factors are considered making the model a better one since it is a nonlinear system. It was found that the most influencing factors of electrical accidents are changing year-on-year basis and the system is unstable. This research is based on reported electrical accidents. Unreported causes and number of electrical accidents are not considered in the study in Regression model. The major cause of accidents is considered in the study for developing the ANN model. Electrical accidents are a major concern and have to be identified and corrected for social wellbeing. The study contributes to the major cause of reported electrical safety issues.



How to Cite

Beena Puthillath, M. Bhasi, C A Babu. (2022). Assessing Reported Electrical Accidents Using Regression and Artificial Neural Network Methods. Computer Integrated Manufacturing Systems, 28(10), 127–146. Retrieved from