Clustering

Question 1

What is the clustering method k-means used for?

Answer 1

number of clusters k is chosen in advance, then the goal is to partition the inputs into sets Si…Sk in a
way that minimizes the total sum of squared distances from each point to the mean of its assigned cluster.

Question 2

How to choose k?

Answer 2

Plotting the sum of squared errors and looking at where the graph “bends”