Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
CI6227 Data Mining – Assignment 3
In this assignment you will get some practical experience working with a clustering algorithm. The gist
of the assignment is to use an implementation of a K-means algorithm and run it to demonstrate the
results of good and bad choices of initial centroids.
Assignment
• Find (or write yourself) a tool that can visualize iterations of K-means algorithm.
• Run it several times
o You can use any 2-dimensional dataset with clearly visible clusters
o You can do any number of clusters ≥ 2
o Take screenshots of the first 5 iterations of every run
o Screenshot should show the data points, the centroids, and the current point-centroid
assignment
• Submit the 5 screenshots from a successful run (centroids reach the true clusters) and from
an unsuccessful run (centroids do not converge into true clusters)
Reporting
Your submission for this assignment is 10 image files (JPG or PNG). Screenshots for the successful run
should be named A1, …, A5 (e.g., A1.jpg, A2.jpg, etc.); screenshots for the unsuccessful run should
be named B1, …, B5.
Submission
Submission should be done in NTULearn. Access the assignment submission page through the left
navigation bar by selecting “Assignments”. Submit ten image files named A1, … A4, A5, B1, B2, …, B5.
Submissions are accepted up to Thursday, December 4th 23:59:59.