k-means clustering algorithm

For the data set shown below, execute the k-means clustering algorithm with k=2 till convergence. You should declare convergence when the cluster assignments for the examples no longer change. As initial values, set µ1 and µ2 equal to x(1) and x(3) respectively. Show your calculations for every iteration. x1 x2 1 1 1,5 2 2 1 2 0,5 4 3 5 4 6 3 6 4

1. You should start your calculation first by initializing your µ1 and µ2 as shown below. µ1 = x(1) =(1,1) µ2 = x(3) =(2,1) 2. For every iteration till convergence find c(i) for i = {1,2,3,4,5,6,7,8} then compute the average for each cluster and reassign the µ1 and µ2 3. Repeat 2 till convergence

[the cyclist]

Read this guide to asking a good question here.

[Image Analyst]

And this: http://www.mathworks.com/matlabcentral/answers/8626-how-do-i-get-help-on-homework-questions-on-matlab-answers

And this: http://www.mathworks.com/matlabcentral/answers/13205-tutorial-how-to-format-your-question-with-markup

And what do you mean by initial values? The kmeans() function doesn't seem to take any initial values.

[the cyclist]

@ImageAnalyst ...

FYI, kmeans does accept a name-value pair ('Start',<value>) for initialization of the cluster centroids.

[Image Analyst]

Thanks for the correction - apparently I overlooked it.

[Walter Roberson]

http://www.mathworks.co.uk/matlabcentral/answers/43073-a-guide-to-tags

Answer 1

[Image Analyst]

Hint:

x1x2 = [...
  1  1
  1.5  2
  2  1
  2  0.5
  4  3
  5  4
  6  3
  6  4]
x1 = x1x2(:, 1);
x2 = x1x2(:, 2);
mu1 = [1,1];
mu2 = [2,1];
for k = 1 : 4
  indexes = kmeans(x1x2, 2, 'start', [mu1;mu2])
  mu1 = mean(x1x2(indexes == 1, :), 1)
  mu2 = mean(x1x2(indexes == 2, :), 1)
end