Skip to main content

Table 3 The r-squared and RMSE values for each type of regression model that was tested to select the best performing model (highlighted in bold) to be the mammalian model

From: EvalDNA: a machine learning-based tool for the comprehensive evaluation of mammalian genome assembly quality

Regression model RMSE R-squared
General linear 16.413 0.775
Elastic net 16.520 0.773
K-nearest neighbors 13.615 0.840
Random forest 12.697 0.860
SVM (linear) 17.190 0.774
SVM (polynomial) 14.363 0.843