Dimensionality Reduction

[lpatiny]

Would be nice to have the following methods

• PCA - ml-pca
• Singular Value Decomposition (SVD) - ml-matrix
• Eigen Decomposition - ml-matrix
• Non-negative Matrix Factorization (NMF)
• tSNE
• Independent Component Analysis (ICA)
• Random Projection
• Factor Analysis (FA)
• umap

[jajoe]

NMF could be interesting because there are a lot of applications (in dimensionality reduction and clustering). I'll implement it.

[Fil]

may I suggest

tSNE

• http://scienceai.github.io/tsne-js/ . (I use this one in my blocks https://bl.ocks.org/Fil/2dc296af097cd26242b52105e06763d8)
• http://cs.stanford.edu/people/karpathy/tsnejs/

umap

• https://github.com/lmcinnes/umap (python only atm)

[jajoe]

Interesting, thanks you for your suggestion

[cannoneyed]

Just wanted to link to a javascript umap implementation here: umap-js. Keep in mind there's a few features missing from the python implementation, but the performance is reasonable all things considered.