CSE 8331 Spring 2008 FINAL PROJECTDUE DATES:Project Choice: 3/31/08 Presentation Slides: 4/21/08 8:00 amInclass Presentations: 4/21-28/08Writeup: 5/10/08 8:00 amGeneral Description:The objective of this project is to give each student the chance to perform some original appliedresearch. The project involves surveying previous work, performing experimental analysiscomparing at least two different algorithms, writing up results, and presenting results in aclassroom presentation.It is hoped that at least a few of these projects will be used as the basis for publications. Requirements:You are to compare the performance of two different algorithms using a dataset of your choice.Each of the algorithms is to be trained and validated using the chosen datasets You decide how totake the existing dataset and divide into training and validation sets. You then need to comparethe performance of the algorithms using two different metrics you have chosen. Grading:Grading for the projects is based on a total of 100 points and make up 40% of each student’sgrade in the course. Grading will be performed on the following basis:Project Identification (10 pts) Due 3/31/08:- Each student is to submit a brief (1 paragraph) writeup of the project topic and howthe experiments will be performed.Presentation Slides (10 pts) Due 4/21/08 8:00am:- Each student is to submit the powerpoint (or other electronic) version of presentation.Presentation (30 pts) 4/21-28/08:- Give a 20 minute presentation in class between 4/21 and 4/28. Your presentationshould provide a summary of your implementation and comparison of your algorithmwith the existing one. Grading is as follows:- Project overview (5pts)- Overall implementation approach & objective (5pts)- Overview of second algorithm (5pts)- Identification of dataset and metrics (5pts)- Implementation overview (tool or language, platform) (5pts)- Preliminary results obtained (5pts)Note that the results at this point in your study may be preliminary. But you musthave implemented enough to present some comparison of the two algorithms.Presentation dates will be assigned based on students’ requests in a FCFS manner.Submission (50 pts) 5/10/08 8:00 am:You are to submit a paper in IEEE format much as you would a submission to aconference. (http://www.engr.smu.edu/~mhd/8331sp08/ieee.doc ). The paper should beabout 10-15 pages in length. Although the actual structure of the paper is up to you, itmust contain the following parts:CSE 8331 1- Problem statement and objectives of paper (5pts)- Related work (5pts)- Detailed description of two algorithms being compared (10pts)- Experimental setup including data, metrics, platform, and experimentsperformed (10pts)- Discussion of results of experiments including figures/graphs (10pts)- Conclusions (5pts)- Bibliography (5pts) (Note that the bibliography should contain some relatedwork needed to do a brief survey for the related works section. Thebibliography should include a minimum of 10 references.)Anyone found plagiarizing at any step in the process will receive a grade of 0 on the project.Students are given the choice of two different projects, and alternative projects will be allowed ifapproved by Professor Dunham. Alternative Project A – Anomaly Detection in Streaming Data:This project requires that you compare an anomaly detection (rare event detection) algorithm ofyour choice against an anomaly detection algorithm implemented using EMM. Recall that EMM(Extensible Markov Model) is a stream modeling technique developed at SMU. Previouspublished studies have shown the effectiveness of EMM in anomaly detection. You are tocompare EMM to an existing anomaly detection algorithm of your choice using data of yourchoice. The EMM algorithm and an anomaly detection algorithm have been developed, and a Web sitewill be made available to you so that you do not have to implement EMM directly. Related Sites:- EMM (http://engr.smu.edu/cse/dbgroup/emm.html )- Rare Event Detection using EMM (http://engr.smu.edu/cse/dbgroup/rare.html )- “Extensible Markov Model” by Margaret H. Dunham, Yu Meng, and Jie Huang, ICDM’04, (http://engr.smu.edu/~mhd/8331sp08/emm.pdf ) - ”Rare Event Detection in a Spatiotemporal Environment,” Yu Meng, Margaret Dunham, Marco Marchetti, and Jie Huang, Proceedings of the IEEE Conference on Granular Computing, May 2006, pp 629-634, , (http://engr.smu.edu/~mhd/8331sp08/rare.pdf ) - A new Web site is being created which will provide access to EMM code. You are to use this site to execute the EMM algorithmAlternative Project B – Biodegradation Prediction:This is an extension of the project performed in CSE 7331 Fall 2007(http://engr.smu.edu/~mhd/7331f07final.html ):The study of biodegradation of compounds in nature is an important research area inEnvironmental Engineering. However, the accurate prediction of which compounds actuallybiodegrade and the speed with which they degrade is a difficult problem yet to be solved.Previous prediction algorithms tend to rely on structural properties of compounds to do theprediction and create somewhat simplistic linear regression models. (Part of the problem withCSE 8331 2previous prediction algorithms is the lack of large amounts of reliable data. Unfortunately thiswill also be a problem with this project.)Your project requires the development and comparison of data mining classification algorithms topredict biodegradability of compounds. A new Web site is being created which will provideaccess to biodegradation data as well as results of previous prediction algorithms. You are to usethe data on this site to implement your classification algorithm and use the included predictionresults for comparison. You may implement any classification/prediction algorithm you choose. CSE 8331