Prediction of the Number of Work Accidents in Coal Mining Sector with Artificial Neural Networks Method for year 2017


Yavuz A., Hacıbektaşoğlu S. E.

International Congress on Occupational Safety and Health, İstanbul, Turkey, 6 - 09 May 2018, pp.545

  • Publication Type: Conference Paper / Summary Text
  • City: İstanbul
  • Country: Turkey
  • Page Numbers: pp.545
  • Istanbul Technical University Affiliated: Yes

Abstract

In our country the coal mining sector is one of the sectors where occupational accidents and occupational diseases are widely experienced. Taking occupational health and safety measures in the coal mining sector is more costly due to the hard work comparing with other sectors. This situation causes employers to stay away from taking occupational health and safety precautions. In this study, it was aimed to predict the number of possible work accidents and mortal work accidents that occurred in 2017 by using work accident data of Health Insurance Agency for 2007-2016. Multilayer artificial neural network is used for this prediction. It is aimed to find the lowest error value of 2017 by changing many parameters in artificial neural network model. In the established network, the data of the year 2016, which is known as the results, are predicted and verified with real values. As a result of the study, it has been seen that not enough importance is given to the safety of work health in the coal mining sector and the negligence of employees is influenced by the number of work accidents. In the last period, dangerous and very dangerous workers are required to have professional qualification certificate and to be able to employ competent staff in these areas. According to the results of the artificial neural network and literature findings we have made, it is predicted that the number of work accidents in the coal mining sector will decrease in 2017.