This paper modifies the loss function of RCA algorithm, gets a robust convergence by restraining the ill con- dition in the iteration of competitive agglomeration among the clusters.
本文通过对RCA算法中遗忘函数的修正,抑制了类间竞争迭代中的病态发散,从而实现了算法的稳健收敛。
In respect of programming, the improved Newton iteration algorithm is associated with function designations, therefore input and transfer of nonlinear functions are solved.
在编程技术方面,将改进的牛顿迭代法与函数调用相结合,解决了非线性函数的输入与转移问题。
By simplifying the objective function of maximum likelihood estimation, the algorithm can realize sequence synchronization and sequence estimation via adaptive iteration and wide window.
通过简化最大似然估计目标函数,提出了用自适应迭代法并结合宽窗口法来实现序列同步和序列估计。
应用推荐