基元抽取问题可以归结为优化问题,即寻找代价函数的全局最小值。
This program can be treated as an optimum program, that is to find the optimum value of a cost function.
许多情况下,该电路可以避免一些局部最小值,而找到一个全局最小值。
In most cases, these circuit structures may avoid some local minima and find a global minimum.
但是由于这些方法不考虑复杂度的控制以及不能保证全局最小值,所以它们的性能不稳定。
However, since these approaches take no account of complexity control and have no guarantee of global minimum, they often suffer from unstable performance.
由于本文提出的角特征不随旋转、平移和尺度的变化而变化,所以配准最终收敛到全局最小值,而不是局部最小值。
Because the proposed angular feature is invariant to rotation, translation and scale transformation, the registration result will converge to global minimization rather than local one.
该算法把惯性权重和学习因子分别通过结合全局和局部最优最小值来进行改写,速度更新公式也做了相应的简化。
The inertia weight and the acceleration coefficient of the algorithm were both adapted by the global best and the local best minimum (GLBM). The velocity equation of the GLBM-PSO was also simplified.
该算法把惯性权重和学习因子分别通过结合全局和局部最优最小值来进行改写,速度更新公式也做了相应的简化。
The inertia weight and the acceleration coefficient of the algorithm were both adapted by the global best and the local best minimum (GLBM). The velocity equation of the GLBM-PSO was also simplified.
应用推荐