KKt

基本解釋

By reformulating the KKT system as a constrained equation, the algorithm generates the search direction by solving a linear equation at each iteration.
通過將問題的KKT系統(tǒng)轉(zhuǎn)化成一個約束方程，算法在每步迭代只需解一個線性方程組即可得到搜索方向。
Under some suitable conditions, theoretical analysis shows that the sequences generated by the method converge globally to the KKT point of the point.
理論分析表明在一些假設(shè)條件下，這一方法產(chǎn)生的序列全局收斂于問題的KKT點。
Methods Learning Motivation Factors Test (KKT) was conducted on 1,186 middle school students.
方法采用學(xué)習(xí)動力因素測驗(KKT)對1186名中學(xué)生測試。
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等結(jié)構(gòu)線性系統(tǒng)的最佳向后擾動分析。
Then we begin to optimize other lagrange multiplications until all multiplications are accord with the KKT term.
我們再對其他Lagrange乘子做最小優(yōu)化，直到所有乘子都符合KKT條件時，目標(biāo)函數(shù)達(dá)到最小，算法結(jié)束。