Machine Learning(Classifier Comparison)

CSE 5160 Course Project

project.html[3/18/21, 9:56:43 AM]

CSE 5160 Machine Learning Course Project Course Project Guidelines Project Suggestions

Classifier Implementation Classifier Comparison Research Papers about Classification

Course Project Guidelines

Your course project is an opportunity for you to explore a machine learning problem of your choice. Below,

you will find some project suggestions.

Your project will be worth 60% of your final class grade, and will have 3 deliverables:

Proposal 10%, 1 page long, due by April 2nd. Project Report 25%, should be maximally 8 pages long, using this report template (adapted from the NeurIPS

conference), due by May 13th Presentation 25%, 10-20 pages slides and 10-15 minutes presentation via Zoom, will be scheduled on April

27, April 29, May 4, May 6, May 11, and May 13.

Please submit the required files to Blackboard by the due date.

There are many datasets out there.

UC Irvine has a repository that could be useful for you project:

http://www.ics.uci.edu/~mlearn/MLRepository.html

https://data.world

Project Suggestions

1. Classifier Implementation

Implement a classifier or classification algorithm. It can be any classification algorithms we learned or have not learned in the class.

Project Proposal

Include the following information:

Project title; The name of classifier or classification algorithm that you will implement;

CSE 5160 Course Project

project.html[3/18/21, 9:56:43 AM]

The programming language you will use; The format of training data you will use, i.e., the source of data set, the attributes' types, the range of attributes'

values, the size of training set, etc.

Project Report

Include the following information:

Abstract The description of implementation process, such as flowchart, functions, pseudocode, or partial codes; The description about 1) using training set to build the classifier, and 2) applying the classifier on a small

amount of test instances. That means, you need to show how your classifier works, and then evaluate your classifier using the training errors and test errors.

Conclusion

Presentation

Explain the source code files; Present how your classifier works; Evaluate your classifier.

2. Classifier Comparison

Compare two or more classifiers or classification algorithms based on a specific data set (containing more than 3000 instances).

Project Proposal

Include the following information:

Project title; The names of classifiers that you want to compare; The data sets that you will use in comparison, including the source of data set, the number of attributes, the

meaning of every attribute, attributes' types, the range of attributes' values, etc; The software you will use in comparison.

Project Report

Include the following information:

Abstract Introduction section: describe the classifiers.

CSE 5160 Course Project

project.html[3/18/21, 9:56:43 AM]

Date set section: describe the data set in detail, and the required data preprocess. Comparison section: describe the parameter setting of classifiers, and show the comparison result on the data

set. Conclusion

Presentation

Uses the slides to present you report, including two classifiers, data set, evaluate and compare the classifiers based on data set.

3. Research Papers about Classifications

Read a research paper published in a journal or conference.

Project Proposal

Include the following information:

The information of research paper you selected, such as title, authors, author affiliation; The journal or conference name that the paper was published; A paragraph about the motivation of the paper, that is, use your own words to restate what problems the

research paper aims to solve.

Project Report

Use your own words to restate the research paper, including introduction, motivation, contribution, and evaluation. The most important thing is to show your understanding to the research conducted in the paper.

Presentation

Uses the slides to summerize the research paper you read, including introduction, motivation, contribution, and evaluation.