Which of the following rationales for applying Principal Components Analysis (PCA) to a given dataset is LEAST likely to be justifiable? Please select a single option below
O It speeds up the training of other algorithms (i.e. algorithms other than the PCA itself) on that dataset
O It reduces the memory needed to store the dataset after a few principal components have been selected
O It is an alternative to using regularization since it effectively reduces the number of features
O It is a useful technique for conducting exploratory data analysis
O All of the above are equally bad justifications for applying PCA