Question

Which among the following is/are some of the assumptions made by the k-means algorithm (assuming Euclidean distance measure)?
A) Clusters are spherical in shape
B) Clusters are of similar sizes
C) Data points in one cluster are well separated from data points of other clusters
D) There is no wide variation in density among the data points

Answer 1

a and b is the correct answer

explanation: The Euclidean distance measure ensures that areas around a cluster centroid comprising points closest to that centroid (which is a cluster) is spherical in shape. Also, this particular distance  measure prevents arbitrarily sized clusters since this typically violates the clustering criterion.