Finding cluster classes in Self Organizing Map.

Hello,

I have training a SOM (Self Organiing Map) using some data with targets in MATLAB2019a. I have generated SOM hits figure and confusion matrix using target matrix. My question is that how do we get to know which cluster of SOM hits figure is representing which class? Is it choosing the same class for a particular instance as I have given in the target matrix? Moreover, confusion matrix is showing very less accuracy. So, it raised one doubt in my mind that it might not be picking the class same as given target matrix. How can we be sure about this?

MATLAB code is as follows:

% input data

x = input';

t = targets;

% Create a Self-Organizing Map

dimension1 = 2;

dimension2 = 2;

net = selforgmap([dimension1 dimension2], 300);

[net,tr] = train(net,x,t);

y = net(x);

classes = vec2ind(y);

view(net)

% Plots

figure, plotsomtop(net)

figure, plotsomnc(net)

figure, plotsomnd(net)

figure, plotsomhits(net,x)

figure, plotsompos(net,x)

figure, plotconfusion(t,y);

Thanks in advance.