In: Statistics and Probability
Use the k-means algorithm and Euclidean distance to cluster the
following eight examples into three
clusters:
A1 = (26, 18), A2 = (20, 26), A3(14, 20), A4(24, 20), A5(14, 30),
A6(22, 18), A7(8, 18), A8(12, 14)
a. Suppose that the initial seeds (centres of each cluster) are A2,
A3, and A8. Run the k-means
algorithm for one epoch only. At the end of this epoch,
o show the new clusters (i.e., the examples belonging to each
cluster);
o show the centres of the new clusters;
o draw a graph space with all eight points, and show the clusters
after the first epoch and
the new centroids.
b. Complete the process for other iterations. How many more
iterations are needed for the
algorithm to converge? Draw the final result of the remaining
epochs.
Note: if an object is at the same minimum distance from two
clusters, consider it as belonging to
both.