Jason H. Moore, Ph.D.I continue to review biomedical #machinelearning papers that do not do HP optimization, do not provide CIs for metrics, misuse CV, don't address data cleaning, don't compare to simpler methods such as regression, don't provide interpretation, etc.... #datascience
Chloé Azencott@moorejh I've been joking for over ten years that I have a template for reviews that already includes "these numbers/figures are difficult to interpret without error bars or confidence intervals" and "a comparaison to simple baselines such as ... and ... would help understand the performance of the algorithm". And yeah, there's so many papers with some form of data leakage out there, it's disheartening.
Veronika Cheplygina@cazencott @moorejh yeah I can also often reuse some kind of "template" for reviews