Ensemble Data Classification based on Diversity of Classifiers Optimized by Genetic Algorithm
| Autor: | Phayung Meesad, Dech Thammasiri |
|---|---|
| Rok vydání: | 2012 |
| Předmět: |
Artificial neural network
Computer science business.industry Data classification General Engineering Decision tree Base (topology) computer.software_genre Machine learning Support vector machine ComputingMethodologies_PATTERNRECOGNITION Genetic algorithm Artificial intelligence Data mining business computer |
| Zdroj: | Advanced Materials Research. :6572-6578 |
| ISSN: | 1662-8985 |
| DOI: | 10.4028/www.scientific.net/amr.433-440.6572 |
| Popis: | In this research we propose an ensemble classification technique base on creating classification from a variety of techniques such as decision trees, support vector machines, neural networks and then choosing optimize the appropriate classifiers by genetic algorithm and also combined by a majority vote in order to increase classification accuracy. From classification accuracy test on Australian Credit, German Credit and Bankruptcy Data, we found that the proposed ensemble classification models selected by genetic algorithm yields highest performance and our algorithms are effective in building ensemble. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|