Monotone Prediction Models in Data Mining

In many decision problems, it is a priori known that the target function should satisfy certain constraints imposed by, for example, economic theory or a human-decision maker. One common type is the monotonicity constraint stating that the greater an input is, the greater the output must be, all other inputs being equal. Well-known examples include investment decisions, medical diagnosis, selection and evaluation tasks. However, often the models obtained by traditional data mining techniques alone does not meet these constraints. Therefore, this book provides a thorough study on the incorporation of monotonicity constraints into a data mining process to improve knowledge discovery and facilitate the decision-making process for end-users by deriving more accurate and plausible decision models. The main contributions include a novel procedure to test the degree of monotonicity of a data set, a greedy algorithm to transform non-monotone into monotone data, and extended and novel approaches to build monotone decision models. The theoretical and empirical findings should be valuable to graduates, researchers and practitioners involved in the study and development of data mining systems.