In reference to the article ‘Data Mining for Fun and Profit’, Predictive models are important in commercial applications where often one is concerned with the possibility of manipulating practices so as to increase sales. In general, of the customer base, 12.5% of them followed the transaction. However, we were able to identify a subgroup of 10% of the customer base that had 43% chance of exhibiting the transaction pattern.
Migliaccio, Giovanni C., Michele Guindani, Su Zhang, and Sudipta Ghorai. "Regression-based Prediction Methods for Adjusting Construction cost Estimates by Project Location." 3rd International/9th Construction Specialty Conference (2011): ...Show more