Skip to main content

Table 5 Random forest models for predicting CIN progression

From: Establishment of multifactor predictive models for the occurrence and progression of cervical intraepithelial neoplasia

NO. Indicators Factors accuracy AUC OOB
1 All clinical features age + menopause+HPV + gravidity+parity+TCT 65.85 67.75 36.59
2 Significant genes TGFBR2 + CSKN1A1 + PRKCI+FOXO1 + CTBP2 73.17 86.75 29.27
3 Significant genes + significant clinical features TGFBR2 + CSKN1A1 + PRKCI+FOXO1 + CTBP2+ menopause+parity+age 75.61 86.25 29.27
4 Genes as the risk factors in unvariable logistic analysis CSKN1A1 + PRKCI+CTBP2 68.29 72.75 24.39
5 Genes as the risk factors in unvariable logistic analysis + Significant genes CSKN1A1 + PRKCI+CTBP2+ menopause+parity+age 68.29 78.75 26.83
6 Genes as the independent factors in multivariable logistic analysis CSKN1A1 + PRKCI 70.73 68.25 21.95
7 Genes as the independent factors in multivariable logistic analysis CSKN1A1 + PRKCI+ menopause+parity+age 68.29 76.75 26.83