Sale!

Project 2: Clustering Algorithms

Original price was: $40.00.Current price is: $35.00.

Category:
Rate this product

Project 2: Clustering Algorithms
Demo time & hard copy report due: October 28
Code submission due: 1pm October 28
Please bring your hard copy report to demo.
Please clearly state the UB Person numbers and UB IT names for all the
group members on the cover of the report.
Two gene datasets (cho and iyer) can be found on Piazza. Please check the README file first for
a short description of the two datasets.
Complete the following tasks:
1. Implement three clustering algorithms to find clusters of genes that exhibit similar expression
profiles: K-means, Hierarchical Agglomerative clustering with Single Link (Min), and one from
(density-based, mixture model, spectral). Compare these three methods and discuss their pros and
cons.
For each of the above tasks, you are required to validate your clustering results using the following
methods:
 Choose an external index (Rand Index or Jaccard Coefficient) and compare the clustering
results from different clustering algorithms. The ground truth clusters are provided in the
datasets.
 Visualize data sets and clustering results by Principal Component Analysis (PCA). You can
implement PCA by yourself or use any existing implementation or package.
2. Set up a single-node Hadoop cluster on your own machine and implement MapReduce Kmeans. Compare with non-parallel K-means on the given datasets. Try to improve the running
time. For single-node Hadoop set up, follow the instructions at setup.pdf which can be found on
Piazza.
Your final submission should include the following:
 Codes: A folder named Code, which contains three clustering algorithms, and MapReduce
K-means algorithm and a README to describe how to run your code.
 Report: A pdf file named Cluster_report .pdf. Describe your implementation details about
all the algorithms. Compare the performance of these approaches using visualization and
external index on the two given data sets. State the pros and cons of each algorithm and any
findings you get from the experiments.
Project Submission:
1. Your final submission should be a zip file named as project2.zip. In the zip file, you
should include aforementioned folder Code and folder Report.
2. Log in any CSE department server and submit your zip file as follows:
>> submit_cse601 project2.zip
The details about Demo will be released on Oct.26 through Piazza. Please note:
 Two new data sets will be given to check your implemented clustering algorithms and
validation measures. The data format is the same with the data sets we already provided.
 During the demo, you will be asked to adopt specific parameter setting and run your code.
Note that copying code/results/report from another group or source is not allowed and may result
in an F in the grades of all the team members. Academic integrity policy can be found at
http://www.cse.buffalo.edu/shared/policies/academic.php

Scroll to Top