cost_alpha_sigma.m

clustering_code ,Clustering Through Ranking On Manifolds Version 0.2 Copyright by Markus Breit

function [cost_cc,clus_close] = cost_alpha_sigma(x_t,dst,num_classes,n,SIGMA_ONLY,dim,lrn_par)

%
% Clustering Through Ranking On Manifolds
% Version 0.2
%
% Copyright by Markus Breitenbach and Gregory Z. Grudic
% This code is for your personal and research use only.
%
% http://www.cs.colorado.edu/~grudic/
% http://ucsu.colorado.edu/~breitenm/
%
% This software is provided "as is," without warranty of any kind, express
% or implied.  In no event shall the authors be held liable
% for any direct, indirect, incidental, special or consequential damages
% arising out of the use of or inability to use this software.
%


if SIGMA_ONLY == 1
    one_over_2_sigma_sq_t = 1/(2*dim* x_t(1)^2);
    my_alpha = 0.99;
else
    
    one_over_2_sigma_sq_t = 1/(2*dim* x_t(1)^2);
    my_alpha = x_t(2);
    if my_alpha > 1
        my_aphal = 1;
    elseif my_alpha < 0
        my_alpha = 0;
    end
    if one_over_2_sigma_sq_t < 0
        one_over_2_sigma_sq_t = 0;
    end
end
    

W = exp(-dst.*one_over_2_sigma_sq_t);

for i = 1:n
    W(i,i) = 0;
end

clear dst;
Dt_s = sum(W');
if min(Dt_s) < eps
    cost_cc = 100;
    return;
end

Dt = Dt_s.^(-1/2);

%Dt = sum(W').^(-1/2);
S = (W.*Dt(ones(n,1),:)).*((Dt(ones(n,1),:))');

iaS = eye(n,n) - (my_alpha * S);

iaS_inv = inv(iaS);
ind_neg = find(iaS_inv<0);
if ~isempty(ind_neg)
    cost_cc = 100;
    return;
end


len_rows = sum(iaS_inv.^2).^(-0.5);
norm_iaS_inv = iaS_inv.*(len_rows(ones(n,1),:));

dot_dist = norm_iaS_inv' * norm_iaS_inv;

mean_dist = mean(dot_dist);

[ind_clust] = Find_Clusters(dot_dist,mean_dist,num_classes,lrn_par.preSet_Centers);

F_norm = norm_iaS_inv(:,ind_clust);

%F = norm_iaS_inv * Y;
%len_col_F = sum(F.^2).^(-0.5);
%F_norm = F.*(len_col_F(ones(n,1),:));
[cost_cc, clus_close] = Cluster_Cost(F_norm,dot_dist,num_classes,n,lrn_par.Cost_function);