Loading
A loss function is the number training tries to make small: cross-entropy for classification (how surprised the model was by the true class) and mean squared error for regression and reconstruction (how far off the predicted values were).
Cross-entropy between two distributions measures the average number of bits needed to encode events from one using a code optimised for the other (Wikipedia); as a loss it penalises confident wrong class probabilities. Mean squared error is the average squared difference between estimates and true values (Wikipedia), the objective for predicting expression, drug sensitivity or reconstructing masked genes. Survival models use the Cox partial likelihood instead, which is why they are not interchangeable with regressors.
Shares Masked autoencoders and masked gene modelling, 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 Masked autoencoders and masked gene modelling, 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 Univariate Cox scores and time-dependent metrics, 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 Masked autoencoders and masked gene modelling, 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.