代码搜索:Regularization
找到约 355 项符合「Regularization」的源代码
代码结果 355
www.eeworm.com/read/161189/10439680
m ttls.m
function [x_k,rho,eta] = ttls(V1,k,s1)
%TTLS Truncated TLS regularization.
%
% [x_k,rho,eta] = ttls(V1,k,s1)
%
% Computes the truncated TLS solution
% x_k = - V1(1:n,k+1:n+1)*pinv(V1(n+1,k+1:
www.eeworm.com/read/418911/10892028
m ttls.m
function [x_k,rho,eta] = ttls(V1,k,s1)
%TTLS Truncated TLS regularization.
%
% [x_k,rho,eta] = ttls(V1,k,s1)
%
% Computes the truncated TLS solution
% x_k = - V1(1:n,k+1:n+1)*pinv(V1(n+1,k+1:n+1))
www.eeworm.com/read/338293/12314657
m ttls.m
function [x_k,rho,eta] = ttls(V1,k,s1)
%TTLS Truncated TLS regularization.
%
% [x_k,rho,eta] = ttls(V1,k,s1)
%
% Computes the truncated TLS solution
% x_k = - V1(1:n,k+1:n+1)*pinv(V1(n+1,k+1:n+1))
www.eeworm.com/read/210916/15189964
m ttls.m
function [x_k,rho,eta] = ttls(V1,k,s1)
%TTLS Truncated TLS regularization.
%
% [x_k,rho,eta] = ttls(V1,k,s1)
%
% Computes the truncated TLS solution
% x_k = - V1(1:n,k+1:n+1)*pinv(V1(n+1,k+1:n+1))
www.eeworm.com/read/367675/2833068
txt 86.txt
发信人: strawman (独上江楼思渺然), 信区: DataMining
标 题: Re: 如何避免"overfitting"
发信站: 南京大学小百合站 (Wed May 28 11:33:03 2003)
“过拟合”问题是个很难避免的问题,我所知道的针对神经网络的方法有两个:
一个是regularization,通过修改performance function来进行。
另