Loading
Logistic regression is the plain linear classifier used as a baseline for subtype prediction; a nearest-centroid classifier instead assigns each sample to the subtype whose average profile it most resembles, which is how PAM50 works.
Logistic regression models the log-odds of a binary outcome as a linear combination of inputs (Wikipedia); a nearest-centroid classifier assigns the label of the class whose mean (centroid) is closest (Wikipedia). Parker and colleagues' PAM50 predictor correlates a sample's fifty-gene profile with five subtype centroids, and using Spearman correlation on ranks makes it robust to platform, so the same classifier runs on microarray, RNA-seq and spatial pseudobulk. Deep models are compared against these because they often do no better.
Shares Spearman rank correlation, Ridge regression and the multilayer perceptron, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Spearman rank correlation, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Ridge regression and the multilayer perceptron, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Linear probe, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares PAM50 / intrinsic subtypes, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Ridge regression and the multilayer perceptron, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Linear probe, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Spearman rank correlation, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.