Use the kmeans() function in the Statistics and Machine Learning Toolbox:
numberOfClasses = 3;
assignedClasses = kmeans(yourData100by5, numberOfClasses);
% Find out which rows are assigned to class #1:
class1Rows = assignedClasses == 1;
% Extract only those observations
class1data = yourData100by5(class1Rows, :); % A 5 column matrix.
% Find out which rows are assigned to class #2:
class2Rows = assignedClasses == 2;
% Extract only those observations
class2data = yourData100by5(class2Rows, :); % A 5 column matrix.
% Find out which rows are assigned to class #3:
class3Rows = assignedClasses == 3;
% Extract only those observations
class3data = yourData100by5(class3Rows, :); % A 5 column matrix.
I don't know how you want to plot a 3 dimensional matrix though. Maybe you can pick just 3 of the 5 features and use scatter3() or plot3().
