This paper makes further discussions onthe constrained learning algorithms (CLA) proposedby Perantonis et al, which is an efficient constrainedback propagation (BP) algorithm formed byimposing the constraint condition...
详细信息
This paper makes further discussions onthe constrained learning algorithms (CLA) proposedby Perantonis et al, which is an efficient constrainedback propagation (BP) algorithm formed byimposing the constraint conditions (most of whichare from the a priori information of the involvedproblems) implicit in the problems on theconventional BP algorithm. In addition, for theproblem of applying CLA's to Finding the roots ofpolynomials based on the relation between the rootsand the coefficients of polynomials, we discuss thecomputation complexities for different number ofconstrained conditions. Finally, some computersimulation results are given to support our claims.
暂无评论