1. Consider any dataset that has more than two class labels. You can create your own or download any publicly available dataset.
(a) Perform K-Means Clustering selecting the best value of k and taking Euclidean distance as similarity measure. Check your algorithm with the following three conditions
i. Maximum number of iterations
ii. Cluster centroid remains unchanged
iii. Highest quality of cluster is reached.
(b) Repeat the Q.2 taking Manhattan distance as similarity measure and note the difference between the clusters as compared to that found in Q. 2.
—– Page 1 of 1 —–
