改进布谷鸟搜索SVM在财务风险评估中的应用

Evaluation model of financial risk based on improved cuckoo search algorithm and support vector machine

针对一种新型智能进化算法——布谷鸟搜索算法提出了基于多群体并行搜索和自适应步长的改进方法。将改进后的方法引入支持向量机参数优化中,提出了基于改进后布谷鸟搜索算法优化支持向量机模型参数的方法并将其应用于上市公司财务风险评估中,有效提高了财务风险模型的分类性能。仿真结果发现:改进的布谷鸟搜索算法用于优化支持向量机参数不仅有效提高了上市公司季度财务数据分类预测精度,而且相较改进之前的布谷鸟搜索算法、遗传算法和粒子群算法具有更快的收敛速度和稳定性。

This study puts forward an improved method based on multiple population parallel search and adaptive search step for a new type of intelligent evolutionary algorithm which is named the cuckoo search algorithm. Then it introduces the improved method to optimize the support vector machine parameters and the detailed program that how to use the improved cuckoo search algorithm to optimize the parameters is proposed and applied to evaluate the listed company＇ s financial risk. The results show that the improved method can effectively improve the classification performance to the quarterly financial data risk model. And the simulation results find that besides the high accuracy, the improved cuckoo search algorithm has faster convergence speed and stability compared to the original cuckoo search algorithm, genetic algorithm and particle swarm optimization.