News for hackers. Stuff that matters.

A comment on ML is useful for many things, but not for predicting scientific replicability. Parent comment

Re:ML is useful for many things, but not for predicting scie... by thwarted on Sunday August 13, 2023 @01:52AM (#37107081)
Our machine learning model used an ensemble of random forest and logistic regression models to predict a paper’s likelihood of replication based on the paper’s text.

This reads like a description of phrenology.

[ Parent | on HN ]
Re:ML is useful for many things, but not for predicting scie... by eimrine on Sunday August 13, 2023 @02:06AM (#37107134)
ML able to predict likelihood of replication, which is also a prediction (about something in Nature) to me is a perfect example of GIGO.
[ Parent | on HN ]
Re:ML is useful for many things, but not for predicting scie... by onigirij on Sunday August 13, 2023 @04:37AM (#37107880)
That is an excellent point/
[ Parent | on HN ]