Bayes与Fisher判别在边坡稳定性中的应用

基于Bayes判别和Fisher判别分析理论并结合工程实际，选用边坡岩土体的重度、黏聚力、边坡角、边坡高度及孔压比5个指标作为边坡稳定性预测的判别因子，建立边坡稳定性预测的Bayes判别和Fisher判别分析模型；以28组边坡已知数据作为学习样本进行训练，建立Bayes和Fisher判别函数及对误判率进行了估计以检验模型的优良性，并选用7个工程实例对所建立的模型进一步进行考察，结果全部预测正确。

Application of Bayes and Fisher Discrimination in Stability of Slope

Such 5 indicators as the gravity, cohesion, slope angle, slope height and pore pressure ratio of the slope rock soil are selected as the discriminant factors predicting the slope stability to establish analysis models of Bayes and Fisher discriminant to predict the slope stability based on Bayes and Fisher discriminant analysis theory and combining with engineering practice. Taking 28 groups of given slope data as the learning sample to train and establish Bayes and Fisher discriminant function, and estimate the false positive rate to test the superiority of the models and select 7 engineering practices to further inspect the established models, all of which predict correctly.