Consider a data set of 1000 observations to be used to train…
Consider a data set of 1000 observations to be used to train and validate a supervised learning algorithm. Compare how these data would be used in a 10-fold cross-validation procedure versus a static holdout procedure in which 70% of the data is used to train the model and 30% of the data is used to validate the model. For both of these approaches, compare: The number of different supervised learning models trained. The size of the training set for each supervised learning model trained. The total number of out-of-sample predictions on which performance measures are computed.