Derivation of Monotone Decision Models from Non-Monotone Data

The objective of data mining is the extraction of knowledge from databases. In practice, one often encounters difficulties with models that are constructed purely by search, without incorporation of knowledge about the domain of application. In economic decision making such as credit loan approval or risk analysis, one often requires models that are monotone with respect to the decision variables involved. If the model is obtained by a blind search through the data, it does mostly not have this property even if the underlying database is monotone. In this paper, we present methods to enforce monotonicity of decision models. We propose measures to express the degree of monotonicity of the data and an algorithm to make data sets monotone. In addition, it is shown that monotone decision trees derived from cleaned data perform better compared to trees derived from raw data