| Autor: |
Mohebpour, M. R., Adznan, B. J., Saripan, M. I. |
| Předmět: |
|
| Zdroj: |
International Review on Computers & Software; Jul2008, Vol. 3 Issue 4, p381-389, 9p, 2 Charts, 4 Graphs |
| Abstrakt: |
In this paper we propose a new method named -Grid Base Classifier- which combines the benefits of the two previous methods in order to improve classification tasks. The problem with the current lazy algorithm is that they learn quickly, but classify very slowly. On the other hand eager algorithms classify quickly but learn slowly. In comparison to these algorithms, the proposed algorithm can both learn and classify quickly. The method is developed based on the grid structure, whereby it was done to create a powerful and robust methods for classification. In the current research, this new algorithm is tested and applied to multiclass classification of two or more categories which are important for handling practical classification problems. The results showed that the new method could improve both the efficiency and accuracy of pattern classification. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
