A Frequency-Based Approach For Multi-Class Data Classification Problem

Data classification is a process of categorizing data into various types or any other distinctive form. In this study, a new frequency-based method is proposed for data classification. This method splits the each attribute to intervals and calculates the frequency of the number of classes occurring in the interval of each attribute. The proposed method is tested on UCI benchmark test problems by 10-fold cross-validation method. The results have shown that the new method reaches to satisfactory solutions in a shorter time compared to the existing results. The better performance is observed in 9 of 11 datasets.