Skip to content

An implementation of [robust version] of Fisher Linear discriminant analysis

Notifications You must be signed in to change notification settings

junlulocky/RobustFisherLDA

Repository files navigation

Robust Fisher Linear Discriminant Analysis (Paper)

This is a collaborative project by Chen LIU and Jun LU.

We use two datasets, both of which are from UCI. They are sonar and ionosphere

In addition, in order to solve the QCQP problem raised by robust LDA, we take advantage of the python library cvxopt and scipt writtern by Jeffery Kline

Code Structure

|-sonar              # the sonar dataset
|-ionosphere         # the ionosphere dataset
|-result             # the result file runs on 10 random samples, used in mainTest.py to show the figure
|-load.py            # load datas
|-FisherLDA.py       # general fisher LDA
|-robustFisherLDA.py # robust fisher LDA 
|-mainTest.py        # test on general fisher LDA and robust fisher LDA
|-log.py  # if you do not want to see so many log infos, please change the DEBUG to False
|-util.py # util functions
|-plotData1.py & plotData2.py       # project the dataset onto 3-d, and plot
|-requirements.txt   # contains the requried packages

About

An implementation of [robust version] of Fisher Linear discriminant analysis

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published