Closed nmondon closed 8 years ago
The weights of the neurons are initialized either to small random values or sampled evenly from the subspace spanned by the two largest principal component eigenvectors
https://en.wikipedia.org/wiki/Principal_component_analysis
https://bl.ocks.org/ktaneishi/9499896
https://www.npmjs.com/package/ml-pca
https://github.com/mljs/pca
The vectors shown are the eigenvectors of the covariance matrix scaled by the square root of the corresponding eigenvalue
https://en.wikipedia.org/wiki/Principal_component_analysis