基于遗传算法的上市公司财务危机预测模型研究

本文以我国沪深A股上市公司为研究对象,选取制造业公司376家,其中被特别处理的ST公司188家,与其配对的健康公司188家,使用遗传算法和21个财务比率建立了财务危机预测模型,并与Logistic回归和BP神经网络模型进行了比较,结果表明,使用遗传算法可以获得不受统计约束且预测准确率更高的模型。

Study on the Predictive Model of Financial Crisis in Chinese Listed Firms Based on the Genetic Algorithm

The author of this paper studied the listed firms of China,established a prediction model of financial crisis by using genetic algorithm and 21 financial ratio,and compared this model with the logistic regression model and artificial neural network model.The research object in this paper is A-share firms in both shenzhen and shanghai capital market 376 Manufacturing companies have been choosed, including 188 Special Treatment companies and 188 normal companies matching with the ST companies. The result showed that the model which predicted more exactly and without statistic limitation could be achieved by genetic algorithm.