<p>This study conducted risk assessment and risk classification on heavy rain damage in the region, then developed the prediction function for heavy rain damage by the risk class. That is to say, the risk index of heavy rain damage by using PSR and DPSIR models was developed for the risk assessment and the risk classes (Red Zone, Orange Zone, Yellow Zone, Green Zone) obtained according to the index. Multiple regression analysis, principal component regression analysis, and artificial neural network(ANN) were applied to develop the prediction function of heavy rain damage. In order to evaluate the prediction performance of the prediction function, we divided heavy rain damage data into the learning section from 2005 to 2012 and the evaluation section from 2013 to 2016. As the results, the ANN using the DPSIR model showed the best prediction performance which has NRMSE of 8.65%. Therefore, the ANN model using the DPSIR was selected as the prediction function in this study. If we can predict the heavy rain damage based on the prediction function, it could be very helpful for disaster preparedness and management.</p><div>
<p>This research was supported by a grant(2018-MOIS31-009) from Fundamental Technology Development Program for Extreme Disaster Response funded by Korean Ministry of Interior and Safety(MOIS).</p>
<div>&#160;</div>
</div>
Development of a Heavy Rain Damage Prediction Function by Risk Classification
