About

In statistics, the training data is the sample whereas in data mining, machine learning, the training data is often a subset of the data set. See Model Building - ReSampling Validation

Training Set represents the hindsight whereas test set represents the foresight.