Copyright @ Wenqi Wang, 2017
The proposed KNN-SC algorithm is KNNG_UOPCA.m in the document. To replicate the reults, cvx is needed to be downloaded from (http://cvxr.com/cvx/) with Mosek (http://cvxr.com/cvx/doc/mosek.html) solver installed.
To run the code
(1) run main_2D.m for 2 dimensional synthetic data
(2) run main_3D.m for 3 dimensional synthetic data
(3) run main_MNIST for MNIST data clustering
(4) run main_Motion for motion data clustering
Note that the comparables in the paper are
(1) Sparse Subspace Clustering (SSC) (http://vision.jhu.edu/code/)
(2) robust subspace clustering via thresholding (TSC) (https://www.nari.ee.ethz.ch/commth/research/downloads/sc.html)
(3) Non-negative Contraint Lasso (NCL), which is implemented using Mosek solver in cvx (http://cvxr.com/cvx/)
When use the code, please cite the following paper:
Unsupervised clustering under the Union of Polyhedral Cones (UOPC) model, Wenqi Wang, Vaneet Aggarwal, and Shuchin Aeron, Pattern Recognitino Letters, Vol 100, pp 104-109 Available: https://www.sciencedirect.com/science/article/pii/S0167865517303653