Suppose we cluster a set of N data points using two different using the k-means clustering algorithm runs but with different number of initial clusters centres. Run 1: 4 initial cluster centres - (a,b), (c,d), (e,f) and (g,h). Run 2: 2 initial cluster centres - (a,b), (c,d) Run 3: 3 initial cluster centres - ((a,b), (c,d), (e,f). Would the clusters created using Run 3 be similar to the 2 or 4 initial cluster runs? Explain your answer briefly.

Suppose we cluster a set of N data points using two different using the k-means clustering
algorithm runs but with different number of initial clusters centres.
Run 1: 4 initial cluster centres - (a,b), (c,d), (e,f) and (g,h).
Run 2: 2 initial cluster centres - (a,b), (c,d)
Run 3: 3 initial cluster centres - ((a,b), (c,d), (e,f).
Would the clusters created using Run 3 be similar to the 2 or 4 initial cluster runs? Explain
your answer briefly.