Investigation of Work Accidents in Turkey with Time Series Analysis and Neural Network Technique

As a result death and permanent absence from work are recorded according to years. Data recorded for several years are modeled by time series analysis. In this study, the result of the number of deaths and the number of permanent absence from work in Turkey, using Box-Jenkins (ARIMA) Technique modeled. The same data are also modeled with Artificial Neural Networks. The models obtained as a result of both methods have been compared in terms of estimation. According to estimates for 2016-2020, the model established with Artificial Neural Networks gave more accurate estimates. According to the ARIMA technique, the number of deaths and number of permanent absence from work in Turkey is expected to decrease for the 2016-2020 period. According to ANN, the number of deaths and number of permanent absence from work in Turkey will increase foreseen for the 2016-2020 period