The algorithm of a decision tree starts with developing a test,
which can best divide groups. The most important objective
of grouping is to obtain a model for prediction. A data set
called ‘training data’, which includes individual variables
and records, are applied for this purpose. In the steps that
follow, the same procedure is performed for lower nodes
with fewer numbers of data in order to obtain the best rules.
Eventually, the tree becomes so large that there remains no