K-means clustering minimizes 'within group sum of squares' to group similar samples, but it does not always converge to the global optimum.
K-means clustering is capable of dividing data into non-overlapping clusters to minimize the 'within group sum of squares.' The objective of k-means is to group similar samples together by minimizing the distance to cluster centers.
One statement that is NOT TRUE about k-means clustering is that it guarantees convergence to the global optimum each time due to the nature of local optima, which can lead to slightly different results upon initialization.
https://brainly.com/question/37817178