10-fold cross validation was chosen to be able to avoid overfitting. Overfitting
is when random error or noise is observed instead of the underlying relationships.
This creates errors in the results. Overfitting often occurs because there is a lack
of training data, usually because the data set is small.