This repository contains the MATLAB code to perform metric learning for a combinatorial datasets. The corresponding paper is listed below:
K.Vaddi, O.Wodo, Metric Learning for High-throughput Combinatorial Data Sets Under review ACS Combinatorial Sciences~(submitted April 2019).
Our software would require MATLAB 2018b or higher with mex files compatibility. Along with the MLCD code, we have also distributed the MT-LMNN code which accompanies the following paper.
Parameswaran, Shibin, and Kilian Q. Weinberger. "Large margin multi-task metric learning." Advances in neural information processing systems. 2010
We thank the authors of MT-LMNN for letting us use the software under MIT-License which is attached to the corresponding folder.
This repository contains the following demos:
- To perform a demo of MT-LMNN from the original paper by Shibin et.al, run
demo_MTLMNN.m
- To perform a demo of MLCD on an open source XRD dataset, run
demo_MLCD.m
We have also provided the codes for the data and analysis used in our paper Vaddi et.al, in the folder MLCD_paper
. You can find more details on reproducing the results from the paper in MLCD_paper
folder.