In this project you will use a selection of pattern classification methods to explore and classify a data set (data sets) of your choice. You will need to let me know about your data selection by April 12 with a project proposal (described below). The data should be challenging (i.e. not linearly separable with an obvious answer) and should require some sort of feature reduction and classifier result verification. You will need to compare the classifiers and assess your feature choice using a number of different analysis approaches including a cross-validation of the data.
Give information about the data set that you plan to use for the project
and what aspects of pattern recognition that you will explore. Examples
include:
| |||||||
Why is this an interesting problem? |
Written report due on Monday May 1 (first day of finals week). | |
20 minute In-class presentation of results during a class period during the final exam period (Monday May 1 at 8AM). A computer will be available so that you can illustrate your results. | |
Matlab code and datafiles needed to run the code. Appendix containing instructions for running the code. |
STATLIB. Datasets from the Statistics Department at CMU. | |
UCI Machine Learning Data Repository |
Page last edited: 04/10/2006