小区域混合大跨度数据分类方程的数学稳定性研究

Small Area Mixed Long-span Data Classification Equation of Stability Study

在数学分类问题中,分类方程的稳定性至关重要,当前的数学分类方程对小区域混合大跨度数据集进行分类的过程中,会产生数据混合和波动问题,存在较大的随机性和偏差,无法确保数据分类方程的稳定性。在原有自调节数学分类方程对小区域混合大跨度数据进行分类的基础上,获取该种数据的分类判决方程,采用非线性微分分析方法获取数据分类方程的常数特解,计算方程的积分曲线走向图,用于加强不同小区域混合大跨度数据分类方程解的稳定性。实验结果说明该种方法对小区域混合大跨度数据进行分类的稳定性控制效率和精度都优于传统分类方程,具有较强的数据分类稳定性,取得了令人满意的效果。

To improve the original classification of mathematical equations, the original self adjustment classification support vector machine （SVM） classification of mathematical equations for small area classified on the basis of the large span data to obtain this data classification decision equation, nonlinear differential analysis method is used to obtain constant special solution of the equation, the calculation equation of integral curve figure, strengthen different small area mixed stability of long-span data classification equations. Experimental results show that the approach of small area mixed long-span data to classify the stability of the control is better than the traditional classification precision and efficiency equation, with strong stability data classification, satisfactory results have been achieved.