为什么不用F1

问题如下：

Exhibit 1 provides results from a sample of loans from the ALPHA model that compared expected and actual defaults over the past 12 months.

While reviewing the model documentation, Lovell confirms that the model was able to correctly predict a default in 5,290 instances of the model prediction dataset after the completed data wrangling.

Based on the results provided for the ALPHA model (Exhibit 1) and its associated documentation, the precision of the model is closest to:

选项：

A.

75.4%

B.

85.5%

C.

95.1%

解释：

C is correct. Precision (P) = TP/(TP + FP), where

TP = True positive = 5,290

FP = False positive = Type 1 error = 273

FN = False negative = Type 2 error = 894

P = 5,290/(5,290 + 273) = 5,290/5,563 = 95.1%.

A is incorrect. The calculation incorrectly uses the total number of predictions in the denominator: 5,290/7,018 = 75.4%.

B is incorrect. The calculation uses the correct equation but incorrectly treats a Type 2 error as a false positive and a Type 1 error as a false negative:

P = TP/(TP + FP) = 5,290/(5,290 + 894) = 5,290/6,184 = 85.5%.