This article is originally published at https://www.tidyverse.org/blog/There is a new release of the embed package on CRAN. embed contains a number of recipe steps that can be used to represent predictors using a smaller set of artificial features. Some of these methods are supervised (i.e., it uses the outcome data) and others are unsupervised (the outcome is not considered). The recipes package already contains similar methods (e.g. principal component analysis, partial least squares, etc.). embed has more sophisticated method and these tend to have more significant package dependencies such as stan and tensorflow.
Thanks for visiting r-craft.org
This article is originally published at https://www.tidyverse.org/blog/
Please visit source website for post related comments.