Skip to main content

Table 9 Comparison between previous studies and this study

From: Application of machine learning techniques for predicting survival in ovarian cancer

Study

Dataset

Model

Class/group

Accuracy

AUC

F1-score

C-index

Explainability

Chen [14]

SEER (4,128 samples)

L2-regularized logistic regression

Binary classification (survived more than 22 months, survived less than 22 months)

0.761

0.621

0.216

-

no

Grimley et al. [15]

SEER

Dataset 1 (39,514 samples)

Ensemble Algorithm for clustering Cancer Data (EACCD)

9 epithelial ovarian carcinoma prognostic groups

-

-

-

0.7391

-

Dataset 2 (25,291 samples, derived from dataset 1)

0.7605

This study

SEER (42,827 samples)

Random Forest

Multiclass classification (5 classes)

0.887

0.823

0.717

-

yes