Scribd
Upload
60 views2 pages

Assignment On Clustering

This document provides instructions for implementing DBSCAN and k-means clustering algorithms on various datasets. Students will first code DBSCAN flexibly to determine optimal parameters from nearest neighbor plots and run it to identify clusters. They will then code k-means, setting k to the number of clusters found by DBSCAN, and run it to cluster the data. Results will include plots and cluster visualizations in different colors, with source code and documents due by the specified deadline.

Uploaded by

Sadat Bin Faruque
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
60 views2 pages

Assignment On Clustering

This document provides instructions for implementing DBSCAN and k-means clustering algorithms on various datasets. Students will first code DBSCAN flexibly to determine optimal parameters from nearest neighbor plots and run it to identify clusters. They will then code k-means, setting k to the number of clusters found by DBSCAN, and run it to cluster the data. Results will include plots and cluster visualizations in different colors, with source code and documents due by the specified deadline.

Uploaded by

Sadat Bin Faruque
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Assignment on Clustering

In this assignment, you will implement DBSCAN and k-means clustering algorithms.

At first, you will have to implement DBSCAN. DBSCAN has two important parameters, EPS and
MINPTS. These are calculated from the plot of (say,) 4-th Nearest Neighbour Distance vs Points
Sorted According to 4th Nearest Neighbour. For the given dataset, you will have to make the plot and
infer appropriate parameters, and use those to run DBSCAN.
You will have to code the DBSCAN with flexible parameters. For some datasets, you may have to use
different EPS and MINPTS. So code rigorously. Using the parameters calculated as above, run
DBSCAN.
Next, you will have to implement the k-means clustering algorithm. For k-means, we will have to
provide the number of clusters to the algorithm (we do not have to do this for DBSCAN). This
number of clusters will be determined from the results of the DBSCAN algorithm.

So in a nutshell, you have to do the following:


1. For each dataset provided, make a plot similar to the following figure, determine DBSCAN
parameters for all the dataset.

2. Run DBSCAN to determine the clusters and number of clusters, show all the clusters in
different colors. The following figure is an example.
Clusters in different colors

3. Run k-means. Set the value of k from the outcome of DBSCAN. If necessary, you can use
bisecting k-means algorithm with proper justification provided in a separate doc file.
Graphically show all the clusters in different colors.
4. Zip source codes, all plot graphs, figures (clustering outcomes) and upload in moodle by 11
am on Saturday, December 12, 2020.

You might also like