Regression modeling involves establishing relationships between variables in a dataset that represent key characteristics. During model training, we begin with a data sample containing features and known label values. This sample is then split into subsets for different stages of the training process. Question: Which subset is best described by the following statement? "A dataset to which we apply an algorithm that determines a function encapsulating the relationship between the feature values and the known label values."