no code implementations • WS 2019 • Victor Ruiz, Lingyun Shi, Wei Quan, Neal Ryan, C Biernesser, ice, David Brent, Rich Tsui
The NB model had the best performance in two additional binary-classification tasks, i. e., no risk vs. flagged risk (any risk level other than no risk) with F1 score 0. 836 and no or low risk vs. urgent risk (moderate or severe risk) with F1 score 0. 736.