Persons working in oil and natural gas activities investigation in terms of work accidents

Occupation accidents occur in many sectors in Turkey every year. In this study, which aims to evaluate occupational accidents occurring in the oil sector, accident estimation was made using artificial neural networks. SPSS 21 package program was used for this purpose. Data on 2210 work accidents that occurred between 2020 and 2023 in a company operating in the oil sector were used. In this study; Artificial neural network modeling was done with monthly accident data. In the study, Multilayer Perceptron Artificial Neural Networks (MLPANN) and Radial Basis Function Artificial Neural Networks (RBFANN), which are feed-forward networks, were used. 70% of the data is divided as training data and 30% as test data. As a result of the analysis; An 84.1% correct classification rate was obtained with the MLPANN method, and an 86.4% correct classification rate was obtained with the RBFANN method. It can be said that the RBFANN method performs more successfully than the MLPANN method. Therefore, it is envisaged that the use of the model in accident risk assessments in the oil industry, which is carried out on a project basis and therefore constantly varies, will be effective for the measures to be taken.