Official LGBT Dadacorn Shirt
The paper reports on negative predictive value and positive predictive value, which are both inherently a measure of the prevalence of the disease in a population. In a study like this where the percentage of positive and negative samples presented to the dogs doesn't reflect the percentage of positive cases in real life, these metrics are meaningless. The only reason to report them here would be if the authors aren't sure what these metrics actually are, or if they're intentionally trying to be misleading. The headline (a news headline, I understand) mentions 94% "accuracy" which is not a meaningful metric is research like this. I can make a machine that detects whether each individual entering an airport is a terrorist or not with 99% accuracy for example.
Buy it: Official LGBT Dadacorn Shirt
Nhận xét
Đăng nhận xét