15 votes
Accepted

Why does Harrell Argue for "Ignoring Y during data reduction"?

All the answers are terrific. I'll just add a "big picture" note. We use data reduction not because it's perfect but when the alternative is a disaster. One of the worst things you can do ...
Frank Harrell's user avatar
9 votes

Why does Harrell Argue for "Ignoring Y during data reduction"?

As far as I understand it, the major reason for not using $Y$ in data reduction is that much theory is based on the crucial assumption that this is not done, and doing it will violate the theory in a ...
Christian Hennig's user avatar
3 votes

Feature selection using backward feature selection in scikit-learn and PCA

First: Doing principal components regression is mostly an alternative to lasso, random forest, and so on, not a preliminary step. Second, lasso is a method of adjusting OLS (or other regressions such ...
Peter Flom's user avatar
  • 117k
2 votes

Finding the most important factor driving the target in a regression problem

Computing variable importance measures without computing confidence intervals for them is highly misleading and ignores the true difficulty of the task. An honest analysis will find that it is almost ...
Frank Harrell's user avatar
1 vote
Accepted

Comparing two groups by the counts of their features

It looks to me like you are trying to analyse RNAseq data, am I wrong ? You should not use t-tests. The distribution of read counts is usually negative binomial. This is why DESeq2 would fit one glm ...
CaroZ's user avatar
  • 594

Only top scored, non community-wiki answers of a minimum length are eligible