Özet:

In the study, different models were developed for interpreting the probabilities of occupational accidents in institutions where severe OSH hazards are present, by taking into account the ANN. The coefficients of the performance determinants (R2) and the Root Mean Square Error of the developed models were evaluated. Models developed with Artificial Neural Networks, training of input and wanted output values by repeatedly applying to the network, learning realized by predicting possible accidents by minimizing faults. In this way, the frequencies of the probable occupational accidents according to the months of the new entries are estimated according to the experiment, and the top management provides reliable decision support. Effective and continuous protection must be provided at a high level in industrial establishments in order to minimize the loss of possible accidents to people and the environment. In this context, institutions are expected to use the “Quantitative Risk Assessment Methodologies” to make decisions with the highest possible accuracy. In the OSH Management Systems, quantitative risk assessments are attempted to reduce the frequencies of possible consequences of the accidents that are scenarios for the hazards. The Artificial Neural Networks (ANN) approach is designed to support the OSH Management System as a Quantitative Risk Assessment. All hazards and risks that may be encountered in Occupational Safety and Health are systematically and efficiently managed, and the process is solved through Occupational Safety and Health Management Systems (OSHMS) within the framework of the continuous improvement approach.

Anahtar Kelimeler: