Prediction the risk of occupational accidents using ANFIS in AZARAB Company

Nowadays, none of the industries are willing to have accidents in their workplaces and use different tools in this regard. One of these tools, which is capable of identifying risks and inappropriate situations, is risk analysis. Due to the importance of job risk prediction and reduction of occupational injury in this study, job risk prediction using different neural network algorithms has been investigated. The purpose of this research is in the field of applied research and in the way of conducting it is in the field of causal and survey research. Accordingly, the test database consists of 119 incidents in past year. Next, due to the high accuracy of the neural network algorithms on the database, it can be concluded that the database has sufficient reliability, in that the Dynamic ANN algorithm has the highest accuracy (76%) in predicting occupational injury. Also, based on the results, the most important criteria affecting the risk of day-time occupational injury, the type of accident and the hazardous situation involved in the accident. This research could have practical application to Azarab, as the company can put all the vulnerabilities together and predict the risk of each of these situations by implementing a neural network algorithm and accordingly offer Take risk control instructions.