Table 3.

Benchmarking analysis of different machine-learning algorithms regarding out-of-sample prediction from white matter microstructure.

	Pearson’s r	Explained Variance Score	Mean Absolute Error	Coefficient of Determination	Mean Squared Error
Ridge Regression	0.2261	0.0562	0.7954	0.0551	0.9394
Random Forest	0.2186	0.0552	0.7949	0.0539	0.9319
Gradient Boosting	0.2202	0.0559	0.7944	0.0546	0.9312
k-Nearest Neighbor	0.1395	0.0099	0.8140	0.0077	0.9775
Kernel Ridge Regression	0.2222	0.0553	0.7927	0.0541	0.9318
Deep Learning	0.1980	0.0406	0.7969	0.0393	0.9464

(cf. Supplementary Table 12)