fcm_SC.m

% File: fcm_SC.m
% Desc: Fuzzy C Means Clustering (Soft Computing Task 1 Extended)
% Date: 01 November 2016

close all;
clear all;
clc;

% Load data train

[dataA,dataB,dataC,dataD,dataE,dataTarget] = loadTrainset;
M = [dataA,dataB,dataC,dataD,dataE];

% Find the centroid from data testing
[centers,U] = fcm(M,8);
maxU = max(U);

% Spare into 8 clusters
index1 = find(U(1,:) == maxU);
index2 = find(U(2,:) == maxU);
index3 = find(U(3,:) == maxU);
index4 = find(U(4,:) == maxU);
index5 = find(U(5,:) == maxU);
index6 = find(U(6,:) == maxU);
index7 = find(U(7,:) == maxU);
index8 = find(U(8,:) == maxU);

% Show the clusters
hold on
plot(M(index1,1),M(index1,2),'ob')
plot(M(index2,1),M(index2,2),'or')
plot(M(index3,1),M(index3,2),'og')
plot(M(index4,1),M(index4,2),'oy')
plot(M(index5,1),M(index5,2),'om')
plot(M(index6,1),M(index6,2),'oc')
plot(M(index7,1),M(index7,2),'ok')
plot(M(index8,1),M(index8,2),'ow')
plot(centers(1,1),centers(1,2),'xb','MarkerSize',15,'LineWidth',3)
plot(centers(2,1),centers(2,2),'xr','MarkerSize',15,'LineWidth',3)
plot(centers(3,1),centers(3,2),'xg','MarkerSize',15,'LineWidth',3)
plot(centers(4,1),centers(4,2),'xy','MarkerSize',15,'LineWidth',3)
plot(centers(5,1),centers(5,2),'xm','MarkerSize',15,'LineWidth',3)
plot(centers(6,1),centers(6,2),'xc','MarkerSize',15,'LineWidth',3)
plot(centers(7,1),centers(7,2),'xk','MarkerSize',15,'LineWidth',3)
plot(centers(8,1),centers(8,2),'xw','MarkerSize',15,'LineWidth',3)
hold off

% Count majority of class from index1
nul1 = 0;
one1 = 0;
for i = 1:size(index1,1)
    if dataTarget(index1(i)) == 0
        nul1 = nul1 + 1;
    else
        one1 = one1 + 1;
    end
end

% Count majority of class from index2
nul2 = 0;
one2 = 0;
for i = 1:size(index2,1)
    if dataTarget(index2(i)) == 0
        nul2 = nul2 + 1;
    else
        one2 = one2 + 1;
    end
end

% Count majority of class from index3
nul3 = 0;
one3 = 0;
for i = 1:size(index3,1)
    if dataTarget(index3(i)) == 0
        nul3 = nul3 + 1;
    else
        one3 = one3 + 1;
    end
end

% Count majority of class from index4
nul4 = 0;
one4 = 0;
for i = 1:size(index4,1)
    if dataTarget(index4(i)) == 0
        nul4 = nul4 + 1;
    else
        one4 = one4 + 1;
    end
end

% Count majority of class from index5
nul5 = 0;
one5 = 0;
for i = 1:size(index5,1)
    if dataTarget(index5(i)) == 0
        nul5 = nul5 + 1;
    else
        one5 = one5 + 1;
    end
end

% Count majority of class from index6
nul6 = 0;
one6 = 0;
for i = 1:size(index6,1)
    if dataTarget(index6(i)) == 0
        nul6 = nul6 + 1;
    else
        one6 = one6 + 1;
    end
end

% Count majority of class from index7
nul7 = 0;
one7 = 0;
for i = 1:size(index7,1)
    if dataTarget(index7(i)) == 0
        nul7 = nul7 + 1;
    else
        one7 = one7 + 1;
    end
end

% Count majority of class from index8
nul8 = 0;
one8 = 0;
for i = 1:size(index8,1)
    if dataTarget(index8(i)) == 0
        nul8 = nul8 + 1;
    else
        one8 = one8 + 1;
    end
end

% Labelling each cluster
label = zeros(8,1,'uint32');
if nul1>=one1
    label(1) = 0;
else label(1) = 1;
end
if nul2>=one2
    label(2) = 0;
else label(2) = 1;
end
if nul3>=one3
    label(3) = 0;
else label(3) = 1;
end
if nul4>=one4
    label(4) = 0;
else label(4) = 1;
end
if nul5>=one5
    label(5) = 0;
else label(5) = 1;
end
if nul6>=one6
    label(6) = 0;
else label(6) = 1;
end
if nul7>=one7
    label(7) = 0;
else label(7) = 1;
end
if nul8>=one8
    label(8) = 0;
else label(8) = 1;
end

% Find the centroid from data testing part 2
[dataA,dataB,dataC,dataD,dataE,dataTarget] = loadTestset;
M = [dataA,dataB,dataC,dataD,dataE];
[centers2,U2] = fcm(M,8);
maxU2 = max(U2);

% Spare into 8 clusters
index1_2 = find(U2(1,:) == maxU2);
index2_2 = find(U2(2,:) == maxU2);
index3_2 = find(U2(3,:) == maxU2);
index4_2 = find(U2(4,:) == maxU2);
index5_2 = find(U2(5,:) == maxU2);
index6_2 = find(U2(6,:) == maxU2);
index7_2 = find(U2(7,:) == maxU2);
index8_2 = find(U2(8,:) == maxU2);

sumin = zeros(8,2,'uint32');
for i = 1:8
    for j = 1:5
        sumin(i,1) = sumin(i,1) + centers(i,j); sumin(i,2) = sumin(i,2) + centers2(i,j);
    end
end

% Find the nearest centroid
kelas = zeros(8,2,'uint32');
x = distfcm(centers,M(i,:));
for i = 1:8
    for j = 1:8
        if (j == 1)
            kelas(i,1) = abs(double(sumin(j,1))-double(sumin(i,2)));
            x = j;
        else
            if kelas(i,1)>abs(double(sumin(j,1))-double(sumin(i,2)))
                kelas(i,1) = abs(double(sumin(j,1))-double(sumin(i,2)));
                x = j;
            end
        end
    end
    kelas(i,2) = label(x);
end

tebakan = zeros(2000,1,'uint32');
for i = 1:size(index1_2,1)
    tebakan(index1_2(i)) = kelas(1,2);
end
for i = 1:size(index2_2,1)
    tebakan(index2_2(i)) = kelas(2,2);
end
for i = 1:size(index3_2,1)
    tebakan(index3_2(i)) = kelas(3,2);
end
for i = 1:size(index4_2,1)
    tebakan(index4_2(i)) = kelas(4,2);
end
for i = 1:size(index5_2,1)
    tebakan(index5_2(i)) = kelas(5,2);
end
for i = 1:size(index6_2,1)
    tebakan(index6_2(i)) = kelas(6,2);
end
for i = 1:size(index7_2,1)
    tebakan(index7_2(i)) = kelas(7,2);
end
for i = 1:size(index8_2,1)
    tebakan(index8_2(i)) = kelas(8,2);
end

indexX = zeros(1,6000,'uint32');

% for i = 1:2000
%     x = distfcm(centers,M(i,:));
% %     if (x(2)>x(1))
% %         indexX(i) = 1;
% %     end
% end

ansTrue = 0;
for i = 1:2000
    if (tebakan(i) == dataTarget(i)) 
        ansTrue = ansTrue + 1;
    end
end
accuracy = ansTrue/20