Skip to main content

Table 2 This table lists five different machine learning algorithms we evaluated for predicting ovarian cancer from chromosome-scale length variation data using the H2O package in R. The algorithms are ranked by the best AUC it achieved using 5-fold cross validation

From: Genetic risk score for ovarian cancer based on chromosomal-scale length variation

Algorithm AUC
Gradient Boosting Machine 0.88
Distributed Random Forest 0.87
Extremely Randomized Trees 0.86
Deep learning 0.82
Generalized Linear Model 0.68