Loading
Domain shift is the mismatch between the data a model was trained on and the data it meets later (cell lines versus tumours, one sequencing platform versus another); domain adaptation is the family of methods that try to bridge it.
Domain adaptation addresses training a model on one data distribution and applying it to a related but different one (Wikipedia). Cell-line screens are the training data for drug response but patients are the goal, so methods such as PRECISE and TRANSACT (Mourragui and colleagues) align the two distributions in a shared subspace and TUGDA (Peres da Silva and colleagues) weights tasks by their uncertainty during adaptation. Platform shift between RNA-seq and microarray is the same problem one level down.
Shares Quantile normalisation, rank transforms and z-scores, Batch effects and harmonisation, External validation, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Quantile normalisation, rank transforms and z-scores, Batch effects and harmonisation, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Out-of-distribution detection (Mahalanobis guard), Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Cell lines as a proxy for patients, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Batch effects and harmonisation, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Out-of-distribution detection (Mahalanobis guard), Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Out-of-distribution detection (Mahalanobis guard), Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares External validation, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.