Why doesn’t my decision tree recognize a simple pattern?
Data Mining algorithms are sometimes perceived to be extremly intelligent and established with many super powers, humans don’t have. In today’s post I’d like to show a simple example in which a mining model fails to detect a very simple rule in the data. This illustrates the need for feature engineering not only to enhance the data with new associations but also to include simple calculations, for example deltas instead of absolute values.
Tags: data mining