引力搜索优化ELM的企业财务危机预警方法

发布时间：2018-03-30 21:05

本文选题：引力搜索算法　切入点：企业危机预警　出处：《计算机应用研究》2017年07期

【摘要】：针对企业财务危机进行了研究,提出一种基于引力搜索算法优化核极限学习机(KELM)的并行模型PHGSA-KELM。模型考虑了特征选择机制和参数优化两者对KELM模型起着同等重要的作用,提出改进的引力搜索算法(HGSA)同步实现特征选择机制和KELM参数优化,同时设计的线性加权多目标函数综合考虑了分类精度和特征子集数量,改善了算法的分类性能,并且基于多核平台的多线程并行方式进一步提高了算法的计算效率。通过真实数据集的实验结果表明,提出的模型不仅获得了较少的特征子集个数,找出了与企业财务危机紧密相关的特征,得到了很高的分类准确率,并且计算效率也得到极大提高,是一种有效的企业财务危机预警模型。
[Abstract]:Based on the study of enterprise financial crisis, a parallel model PHGSA-KELM-based on Gravity search algorithm to optimize kernel limit learning machine (KELM) is proposed. The model considers the feature selection mechanism and parameter optimization, which plays an equally important role in the KELM model. An improved gravity search algorithm (HGSA) is proposed to synchronize feature selection mechanism and KELM parameter optimization. The linear weighted multiobjective function is designed to consider the classification accuracy and the number of feature subsets in order to improve the classification performance of the algorithm. The multi-thread parallel method based on multi-core platform further improves the computational efficiency of the algorithm. The experimental results of real data sets show that the proposed model not only obtains a small number of feature subsets. This paper finds out the characteristics closely related to the enterprise financial crisis, obtains a very high classification accuracy, and greatly improves the computational efficiency. It is an effective early warning model of enterprise financial crisis.
【作者单位】：深圳信息职业技术学院数字媒体学院;
【基金】：国家自然科学基金青年基金资助项目(61303113) 广东省自然科学基金资助项目(2016A030310072)
【分类号】：F275;TP18

【相似文献】