Table 3 Model performance in discrimination according to the cutoff predicted FEV₁/FVC ratio in the test set

60.5%

(52.2–68.5)

83.1%

(80.8–85.3)

31.9%

(26.5–37.7)

94.2%

(92.5–95.5)

71.8%

(66.5–76.9)

73.5%

(65.6–80.4)

75.4%

(72.8–77.9)

28.1%

(23.6–32.8)

95.6%

(94–96.9)

74.4%

(69.2–79.1)

82.3%

(75.2–88.1)

68.6%

(65.8–71.3)

25.5%

(21.7–29.7)

96.7%

(95.3–97.9)

75.5%

(70.5–79.7)

87.8%

(81.3–92.6)

60.6%

(57.7–63.5)

22.6%

(19.2–26.2)

97.4%

(96–98.5)

74.2%

(69.5–78)

95.9%

(91.3–98.5)

44.9%

(420–47.8)

18.5%

(15.8–21.5)

98.8%

(97.5–99.6)

70.4%

(66.6–73.2)