Nice writeup. F1, balanced accuracy, etc. In truth it depends on your problem and what a practical "best" solution is, especially in imbalanced scenarios, but M

Nice writeup. F1, balanced accuracy, etc. In truth it depends on your problem and what a practical "best" solution is, especially in imbalanced scenarios, but Matthews Correlation Coefficient (MCC) is probably the best comprehensive and balanced blind go-to metric, because it guarantees that more portions of the confusion matrix are good [0,1]. I made a quick interactive, graphical exploration to demonstrate this in python [2]. [0]: https://biodatamining.biomedcentral.com/articles/10.1186/s13040-023-00322-4 https://biodatamining.biomedcentral.com/articles/10.1186/s13... [1]: https://biodatamining.biomedcentral.com/articles/10.1186/s13040-021-00244-z https://biodatamining.biomedcentral.com/articles/10.1186/s13... [2]: https://www.glidergrid.xyz/post-archive/understanding-the-roc-curve-and-beyond https://www.glidergrid.xyz/post-archive/understanding-the-ro...