论文部分内容阅读
加权线性支持向量分类机是数据挖掘的新方法.它对应于一个优化问题.针对加权线性支持向量分类机优化问题建立了数据扰动分析理论方法.具体地针对加权线性支持向量分类机的原始问题建立了数据扰动分析基本定理,定理可以得到加权线性支持向量分类机问题的解及决策函数对数据参数的偏导数,同时可以定量分析输入数据的误差以及数据各种变化对其解以及决策函数值的定量影响,可以回答加权线性支持向量分类机问题的稳定性问题和灵敏度分析问题.
Weighted linear support vector classification is a new method of data mining which corresponds to an optimization problem.Aiming at the optimization problem of weighted linear support vector classification machine, a theoretical method of data perturbation analysis is established.For the original problem of weighted linear support vector classification machine The basic theorem of data perturbation analysis theorem can get the solution of the weighted linear support vector machine and the partial derivative of the decision function to the data parameter. At the same time, we can quantitatively analyze the errors of the input data and the changes of the data to their solutions and decision function values Quantitatively, we can answer the questions of stability and sensitivity of weighted linear support vector machines.