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。
The elicitation search and integrated forward direction, reverse direction and bidirectional reasoning was adopted to set up expert system in reasoning machine design.
推理机设计采用启发式搜索，结合正向、反向及双向推理建立专家系统。
The elicitation search and integrated forward direction, reverse direction and bidirectional reasoning was adopted to set up expert system in reasoning machine design.
推理机设计采用启发式搜索，结合正向、反向及双向推理建立专家系统。
应用推荐