KKt

基本解释

By reformulating the KKT system as a constrained equation, the algorithm generates the search direction by solving a linear equation at each iteration.
通过将问题的KKT系统转化成一个约束方程，算法在每步迭代只需解一个线性方程组即可得到搜索方向。
Under some suitable conditions, theoretical analysis shows that the sequences generated by the method converge globally to the KKT point of the point.
理论分析表明在一些假设条件下，这一方法产生的序列全局收敛于问题的KKT点。
Methods Learning Motivation Factors Test (KKT) was conducted on 1,186 middle school students.
方法采用学习动力因素测验(KKT)对1186名中学生测试。
The techniques, described here, are suitable to other systems such as KKT and SQD.
所用方法亦可用于研究KKT及SQD等结构线性系统的最佳向后扰动分析。
Then we begin to optimize other lagrange multiplications until all multiplications are accord with the KKT term.
我们再对其他Lagrange乘子做最小优化，直到所有乘子都符合KKT条件时，目标函数达到最小，算法结束。