Constructive Induction of Cartesian Product Attributes

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

AutoML: A Survey of the State-of-the-Art (2019, neutral): “product [84] and M-of-N [85] are commonly used for nominal features.”
Bayesian Variable and Transformation Selection in Linear Regression (2002, background): “Much of this literature concerns itself with improving predictiveperformance throughthe inclusionof new predictors that are functionsof multiple existing predictors; see, for example, Pazzani (1996).”
Not So Naive Bayes: Aggregating One-Dependence Estimators (2005, methodology): “Notwithstanding the fact that some violations of the attribute independence assumption do not matter (Domingos & Pazzani, 1996), it is clear that many do, and there is an increasing body of work developing techniques to retain NB’s desirable simplicity and efficiency while alleviating the problems…”
Locally Weighted Naive Bayes (2002, methodology): “Both semi-naive Bayes (Kononenko, 1991) and the Cartesian product method (Pazzani, 1996) are such transformation-based attempts for capturing pairwise dependencies.”