Table 2 Algorithms used in this work and their respective hyperparameter optimization spaces
From: The effect of noise on the predictive limit of QSAR models
Algorithm | Hyperparameters searched in optimizationa,b |
|---|---|
Ridge regression (Ridge) | PCA n components \(\in \left( {1,{ }3,{ } \ldots { },59} \right)\) α \(\in \left( {1, 2, 3, 4, 5, 10} \right)\) |
k-nearest neighbors (kNN) | PCA n components \(\in \left( {1,{ }3,{ } \ldots { },59} \right)\) k \(\in \left( {1,{ }2,{ } \ldots ,{ }20} \right)\) |
Support vector regressor (SVR) | PCA n components \(\in \left( {1,{ }3,{ } \ldots { },59} \right)\) C \(\in \left( {0.01,{ }0.1,{ }1,{ }10} \right)\) kernel: radial basis function (RBF) |
Random forest (RF) | PCA n components \(\in \left( {1,{ }3,{ } \ldots { },59} \right)\) n estimators \(\in \left( {1,{ }10,{ } \ldots ,{ }200} \right)\) max depth \(\in \left( {1,{ }3,{ } \ldots ,{ }99} \right)\) max leaf nodes \(\in \left( {2,{ }12,{ } \ldots ,{ }92} \right)\) |
Gaussian process (GP) | PCA n components \(\in \left( {1,{ }3,{ } \ldots { },59} \right)\) kernel:c RBF, WhiteKernel, Matern, DotProduct, ExpSineSquared, ConstantKernel or RationalQuadratic Normalize y: true |