Hybrid learning methods use theoretical knowledge of a domain and a set of classified examples to develop a method for classification. Methods that use domain knowledge have been ...
Multi-class problem is the class of problems having more than one classes in the data set. Bayesian Network (BN) is a well-known algorithm handling the multi-class problem and is ...
There is an overwhelming increase in submissions to genomic databases, posing a problem for database maintenance, especially regarding annotation of fields left blank during submi...
Gustavo E. A. P. A. Batista, Maria Carolina Monard...
We present our experience in applying a rule induction technique to an extremely imbalanced pharmaceutical data set. We focus on using a variety of performance measures to evaluate...
This article, which lies within the data mining framework, proposes a method to build classifiers based on the evolution of rules. The method, named REC (Rule Evolution for Classif...