1.

The K-means algorithm automatically finds the correct number of clusters for a given dataset.

true or false

2.

What is one reason for running the K-means algorithm multiple times using different initial cluster centroids?

to find a fast converging solution

to avoid local minima in the solution

to find the best norm

to account for possible errors in computing distances

to avoid the fire swamp

What is one reason for running the K-means algorithm multiple times using different initial cluster centroids?