Learning-test split method

Split sample is a popular resampling method, also referred to as the learning-test split or holdout method (McLachlan, 1992), which entails a single partition of the observed data into a training set and a test set. The size of the two sets is based on a predetermined proportion p for the test set. For example, if p =, this method allots two-thirds of the observed data to the training set and one-third to the test set. Figure 10.5 illustrates this resampling method. 

Most current studies report an internal validation accuracy without an independent validation set. When there are a sufficiently large number of samples, the whole dataset can be split into two, one for training and one for testing (validation) this method is called hold-out validation. When the number of samples is limited, leave-one-out cross validation (LOOCV) is a popular technique. Here, a procedure is repeated N times, and each time a different sample is left out and used for testing the model learned from the remaining (N - 1) samples. The accuracy of... 

---