另外,自然编码方式与期望值选择机制也提高了算法的执行效率。
In addition nature coding and expected value selection also improve the efficiency of GA.
应用改进后的基于一阶微分期望值的亚像素边缘检测算法,可以快速、精确地检测到边缘的位置。
Using improved sub-pixel edge detecting algorithm based on the expectation of first-order derivatives, we can fast and precisely detect the edge position.
获得的值与对应的期望值对比,误差使用反传学习算法传回网络,用以更新连接权值。
Compared obtained value with corresponding expected value, error was sent to network with reverse learning algorithm, so that renovate connect weighting value.
应用推荐