This is a machine learning related question:

Some authors present a machine learning/data mining pipeline process with only 3 main phases instead of those 6 shown in the image below (see the arrows). Explain the reasons why they probably decided to do that including in your answer what are those 3 main phases and their corresponding relevance to build knowledge. Evaluation Machine Learning Data Mining Knowledge L.O. Transformation Patterns Processing Transformed data Selection Processed Data Target data Data