Flood disaster loss comprehensive evaluation model based on optimization support vector machine

As an important part of flood disaster risk-control system, flood disaster evaluation has characteristics like complex and uncertain relationships between index of flood hazard evaluation and grade of flood hazard evaluation and few measured sample etc. Therefore, a support vector machine model, which is fit for grade of evaluating flood disaster and optimized support vector machine, is established to realize the function that can comprehensively evaluate index of multi-dimensional disaster situation in one-dimensional continuous space. The model effectively solves problems like the one that grade of every single index is incompatible with one another so as to improve the accuracy of disaster evaluation results. Besides, generality of the model is improved by adopting different grade standard for disaster situation index in the relative regions. The results show that the model based on optimization support vector machine has a strong generalization and evaluation results is reasonable so that the model is considered to have a good application prospect in multi-index comprehensively evaluation. (C) 2009 Elsevier Ltd. All rights reserved.