earticle

영어

The construction industry has seen a continuous increase in the number of victims over the last five years, despite various efforts and investments in disaster prevention. This means that we need substantial and fundamental preventive measures that are different from the ones we have now. Therefore, this study proposed a prediction model that can predict construction disasters in advance using a highly predictable random forest method. The random forest technique was used to analyze the impact factors of the construction disaster, and the model was constructed by selecting the factors of high importance from the variables. As a result of comparing the accuracy of the random forest model and the discrimination analysis model in order to verify the constructed construction disaster prediction model, it is found that the random forest model is more accurate than the discrimination analysis model. Accordingly, it is expected that disaster types that can occur by the prediction model can be accurately predicted in advance and preferentially managed to contribute to disaster prevention.