![]() | Only 14 pages are availabe for public view |
Abstract Classification problems can be seen in many real-life applications, such as computational biology, communications, and the network security. The main aim in such problems is to build a classifier that finds mathematical or logic rules to assign each sample in the training data to the correct label. Then, these rules can be used to assign labels for samples in the testing data. Neural Networks, Support Vector Machines, Induction Decision Trees and Random Forest methods have been used extensively in classification problems. Nevertheless, these methods have some drawbacks in terms of the accuracy and/or the structure of solutions. Therefore, there is a need to adopt new methods to generate more efficient logical rules and overcome drawbacks of these methods. |