yinjiehuang / r2lml Goto Github PK

Yinjie Huang

University of Central Florida

2013

###CONTENTS###

• General Information
• Requirements
• Installation
• Usage
• Citation

=============================================

###GENERAL INFORMATION###

This software was written to implement the algorithm "Reduced-Rank Local Distance Metric Learning". All rights belong to the author.

###REQUIREMENTS###

To run this software, you need to have the following components installed:

• Mathworks MATLAB

###INSTALLATION###

This software doesn't require any installation. Just drop the files into a folder.

###USAGE###

• Please put the data set in the folder 'Data'. The data format is '<Dimension+1> X <# of data>'. The last row is the label of the data.

• To run the software, run the file 'Demo.m' in which many parameters could be set. Please refer to the paper how to setup these parameters.

• If you want to save the testing result label, please go to '\Functions\R2LML'. Uncomment line 81. The label will be saved automatically.

###Citation###

Please cite the following paper:

• Yinjie Huang, Cong Li, Michael Georgiopoulos and Georgios C Anagnostopoulos, 'Reduced-Rank Local Distance Metric Learning', ECML/PKDD 2013.