It looks like you're new here. If you want to get involved, click one of these buttons!
Hi, I was wondering if you can explain or point me to what are the priors used in the H-measure (in Binary Logistic Regression, classification performance metrics)?
Can you post a screenshot, so I know exactly what you are referring to?
Hi EJ, sorry for losing track of this post, see the red circles in the enclosed figure (it is in the binary logistic regression)
Ah, this is an analysis I am not expert on. But I do know it is a frequentist analysis. So is the H-measure a Bayesian concept? I will ask Erik-Jan (no, this is not strange).
the H-measure performance metric is "just another" measure of classification performance. For more information about this H-measure you can read the paper by its developer David Hand. For a slightly more approachable introduction I suggest the website https://www.hmeasure.net/.
Hope this helps!
Thank you Erik-Jan, yes I'm just trying to understand how exactly the weights are assigned to the TP/TN versus FP/FN (for, the standard accuracy measure clearly is very biased for unbalanced classes, as is often the case in medicine..).
Thank you for the link!
EJ > I've had a could very "bayesian" students in my lab for a few years, and now that JASP exists, well, that's all they talk about (other than "how big" their Bayes Factor is.. :-) :-) :-) )
We use the default implementation in the package hmeasure on CRAN. By default the severity ratio is set to be reciprocal of relative class frequency. You can see further details in the HMeasure section in the package documentation. https://cran.r-project.org/web/packages/hmeasure/hmeasure.pdf