Please read the instructions carefully.
Implement the following algorithms from scratch. You may use any programming
language that you like.
- Single linkage
- Complete linkage
- Average linkage
- Lloyd’s method
Test it on 2 to 3 data sets. You may use any data sets. For example. The UCI datasets
page includes any excellent ones: [login to view URL] Please
choose the data carefully, so that it is appropriate for clustering. (Ex. the Iris data set).
In addition, it is highly advised to test your algorithm on some of the small examples
like the ones we did in class, in order to make sure that the implementation is correct.
Compare the output of your algorithm against the “target truth” in the data sets
you use, by using Hamming distance. Summarize your findings in a table, and
specify which algorithm did best for each data set.
The above can earn you a grade of up to a B. To reach up to a grade of A+, you need
to complete one of the following:
- K-means++ ([login to view URL])
- BDSCAN clustering algorithm
- Any other clustering algorithm
- Compare clusterings using misclassification error (in addition to Hamming distance):
[login to view URL]
(Section 5)
Hi, I'm working in machine learning field for last 1 year and I really like exploring this field as much as possible. I've worked on various projects and participated in may hackathons.
I'd like to take on this project and I can assure you I'll complete this.
Hi we are expert in AI and ML including Supervised , un-supervised ML, Neural network and Natural language processing using Keras , Tenserflow and Theano