Heavy-tailed Symmetric Stochastic Neighbor Embedding (HSSNE)

Click here to download the toolbox, which contains the following files:

fphssne.m	HSSNE for multi-dimensional input using the fixed-point algorithm.
fphssne_p.m	HSSNE for pairwise similarities input using the fixed-point algorithm.
x2p.m	Utility function that converts multi-dimensional input to pairwise similarities.
GetVisualizationMovie.m	Utility function that converts intermediate visualization maps to a MATLAB movie.
dist2q.m	Utility function that computes the heavy-tailed embedding similarities for given pairwise squared distances.
distSqrdSelf.m	Utility function that computes the pairwise squared distances in a fast manner.
fphssne_demo.m	Demonstration script.
iris.mat	A small dataset in MATLAB format used by the demo.

Please give reference to our contribution below if you use our HSSNE toolbox in your publishable work.

Zhirong Yang, Irwin King, Zenglin Xu, and Erkki Oja. Heavy-Tailed Symmetric Stochastic Neighbor Embedding. In Advances in Neural Information Processing Systems 22 (NIPS2009), pages 2169-2177, Vancouver, Canada, 2009. [pdf]