Table 3 Performances of different models in predicting lymph node metastasis in patients with cervical cancer
Models | Cutoff | Sensitivity (95%CI) | Specificity (95%CI) | PPV (95%CI) | NPV (95%CI) | AUC (95%CI) | Accuracy (95%CI) |
|---|---|---|---|---|---|---|---|
XGBoost model | |||||||
Training set | 0.413 | 0.886(0.780–0.991) | 0.890(0.826–0.954) | 0.756(0.625–0.888) | 0.953(0.908–0.998) | 0.939(0.938–0.940) | 0.889(0.834–0.944) |
Testing set | 0.413 | 0.786(0.571-1.000) | 0.600(0.448–0.752) | 0.407(0.222–0.593) | 0.889(0.770-1.000) | 0.721(0.716–0.727) | 0.648(0.521–0.776) |
Logistic Regression model | |||||||
Training set | 0.263 | 0.657(0.500-0.814) | 0.670(0.574–0.767) | 0.434(0.301–0.567) | 0.836(0.751–0.921) | 0.687(0.683–0.690) | 0.667(0.584–0.749) |
Testing set | 0.263 | 1.000(1.000–1.000) | 0.150(0.039–0.261) | 0.292(0.163–0.420) | 1.000(1.000–1.000) | 0.812(0.809–0.816) | 0.370(0.242–0.499) |
MNB model | |||||||
Training set | 0.508 | 0.371(0.211–0.532) | 0.901(0.840–0.962) | 0.591(0.385–0.796) | 0.788(0.710–0.867) | 0.611(0.607–0.615) | 0.754(0.679–0.829) |
Testing set | 0.508 | 0.429(0.169–0.688) | 0.900(0.807–0.993) | 0.600(0.296–0.904) | 0.818(0.704–0.932) | 0.745(0.740–0.750) | 0.778(0.667–0.889) |
SVM model | |||||||
Training set | 0.094 | 0.829(0.704–0.953) | 0.824(0.746–0.902) | 0.644(0.505–0.784) | 0.926(0.869–0.983) | 0.830(0.827–0.832) | 0.825(0.759–0.892) |
Testing set | 0.094 | 0.929(0.794-1.000) | 0.100(0.007–0.193) | 0.265(0.142–0.389) | 0.800(0.449-1.000) | 0.696(0.690–0.703) | 0.315(0.191–0.439) |
Decision Tree model | |||||||
Training set | 0.668 | 0.886(0.780–0.991) | 0.440(0.338–0.542) | 0.378(0.273–0.483) | 0.909(0.824–0.994) | 0.691(0.688–0.694) | 0.563(0.477–0.650) |
Testing set | 0.668 | 0.643(0.392–0.894) | 0.775(0.646–0.904) | 0.500(0.269–0.731) | 0.861(0.748–0.974) | 0.724(0.719–0.729) | 0.741(0.624–0.858) |
Random Forest model | |||||||
Training set | 0.251 | 0.829(0.704–0.953) | 0.758(0.670–0.846) | 0.569(0.433–0.705) | 0.920(0.859–0.981) | 0.875(0.873–0.877) | 0.778(0.705–0.850) |
Testing set | 0.251 | 0.643(0.392–0.894) | 0.650(0.502–0.798) | 0.391(0.192–0.591) | 0.839(0.709–0.968) | 0.684(0.678–0.690) | 0.648(0.521–0.776) |
GBDT model | |||||||
Training set | 0.273 | 0.971(0.916-1.000) | 0.978(0.948-1.000) | 0.944(0.870-1.000) | 0.989(0.967-1.000) | 0.997(0.997–0.997) | 0.976(0.950-1.000) |
Testing set | 0.273 | 0.429(0.169–0.688) | 0.750(0.616–0.884) | 0.375(0.138–0.612) | 0.789(0.660–0.919) | 0.651(0.645–0.657) | 0.667(0.541–0.792) |