Cross-chromophore test

Definition: Cross-chromophore test is an evaluation strategy for prediction models that builds on the ability of NIRS to measure multiple chromophores. As a type of out-of-distribution test, it tests for generalization across chromophores by reserving all records of one chromophore for training while using the other chromophore exclusively to quantify the model’s predictive capacities.

Alternative definition:

Synonym:

References:

Related terms: