inst/gallery/Pulsar/README.md

Pulsar data (HTRU2) from UCI Machine Learning Repository. Goal is to predict pulsar or non-pulsar star.

All plots give good separation of red and black. But look at the one using prVis() at degree 3! Even without labels, one would see that "something is happening."

This also shows that one should get almost perfect prediction using a polynomial regression model (say logistic) of degree 3.

tsne with labels umap with labels

prvis with labels prvis without labels prvis with labels



matloff/prVis documentation built on Dec. 3, 2019, 9:27 a.m.