An R package that implements 11 distance metric learning methods. The implemented methods are:
1- Free Energy Nearest Neighbor (FENN)
2- Free Energy Nearest Neighbor with dimension reduction (subFENN)
3- Discriminant Adaptive Nearest Neighbor (DANN)
4- iterative Discriminant Adaptive Nearest Neighbor (iDANN)
5- Locally Adaptive Metric Nearest Neighbor (ADAMENN)
6- Locally Adaptive Metric Nearest Neighbor (iADAMENN)
7- Xing's method
8- Relevant Component Analaysis (RCA)
9- Linear Fisher Discriminant Analaysis (LFDA)
10- Discriminant Component Analaysis (DCA)
11- Neighborhood Component Analysis (NCA)
All methods are implemented using RcppArmadillo
in order to attain improved performance. A vignette detailing the usage of each method is attached to the package.
You can install this R pacakge using the following:
library(devtools)
install_github("carltonyfakhry/DistanceLearning")
Please see the Vignette for this package using the following:
browseVignettes("DistanceLearning")