经典流形学习 ISOMAP算法程序中的一个问题
读程序时,有一点看不懂,牛人来瞧瞧,指点指点啊。
在看流形学习的一段经典代码时,遇到的一个问题。 函数如下:
function [Y, R, E] = IsomapII(D, n_fcn, n_size, options);
% ISOMAPII Computes Isomap embedding using an advanced version of
% the algorithm in Tenenbaum, de Silva, and Langford (2000),
% which can take advantage of sparsity in the graph and
% redundancy in the distances.
%
% [Y, R, E] = isomapII(D, n_fcn, n_size, options);
%
% Input:
% D = input-space distances between pairs of N points, which can
% take 1 of 3 forms:
% (1) a full N x N matrix (as in isomap.m)
% (2) a sparse N x N matrix (missing entries are treated as INF)
% (3) the name of a function (e.g. 'd_fun') that takes
% one argument, i, and returns a row vector containng the
% distances from all N points to point i.
%
% n_fcn = neighborhood function ('epsilon' or 'k')
% n_size = neighborhood size (value for epsilon or k)
代码中,提示 n_fcn 是近邻函数,后面的'epsilon' 或'k' 是什么意思,是指函数的类型吗
整个函数开始要求输入参数,这个n_fcn是什么形式,
牛人多帮帮,小弟先谢过了。