New Dynamic Classifiers Selection Approach for Handwritten Recognition
In this paper a new approach based on dynamic selection of ensembles of classifiers is discussed to improve handwritten recognition system. For pattern classification, dynamic ensemble learning methods explore the use of different classifiers for different samples, therefore, may get better generalization ability than static ensemble learning methods. Our proposed DECS-LR algorithm (Dynamic Ensemble of Classifiers Selection by Local Reliability) enriched the selection criterion by incorporating a new Local-Reliability measure and chooses the most confident ensemble of classifiers to label each test sample dynamically. Confidence level is estimated by proposed reliability measure using confusion matrix constructed during training level. After validation with voting and weighted voting fusion methods, ten different classifiers and three benchmarks, we show experimentally that choosing classifiers ensemble dynamically taking into account the proposed L-Reliability measure leads to increase recognition rate for Handwritten recognition system using three benchmarks.
KeywordsMultiple classifier system Dynamic classifier selection Local accuracy estimation Classifiers fusion Handwritten recognition
Unable to display preview. Download preview PDF.
- 3.Kuncheva, L.I., Whitaker, C.S.C., Duin, R.P.W.: Is independence good for combining classiers. In: Proceedings of the 15th International Conference on Pattern Recognition, Barcelona, Spain, pp. 168–171 (2000)Google Scholar
- 5.Azizi, N., Farah, N., Khadir, M., Sellami, M.: Arabic Handwritten Word Recognitionv Using Classifiers Selection and features Extraction / Selection. In: 17 th IEEE Conference in Intelligent Information System, IIS 2009, Poland, pp. 735–742 (2009)Google Scholar
- 8.Azizi, N., Farah, N., Sellami, M.: Off-line handwritten word recognition using ensemble of classifier selection and features fusion. Journal of Theoretical and Applied Information Technology, JATIT 14(2), 141–150 (2010)Google Scholar
- 9.Azizi, N., Farah, N., Sellami, M.: Ensemble classifier construction for Arabic handwritten recognition. In: The 7th IEEE International Workshop in Signal Processing and Sustems, WOSSPA 2011, Tipaza, Algeria, May 8-10 (2011)Google Scholar
- 10.Azizi, N., Farah, N., Sellami, M.: Progressive Algorithm for Classifier Ensemble Construction Based on Diversity in Overproduce and Select Paradigm: Application to the Arabic handwritten Recognition. In: The 2nd ICICS 2011, Jordan, May 22-24, pp. 27–33 (2011)Google Scholar
- 11.Farah, N., Souici, L., Sellami, M.: Classifiers combination and syntax analysis for arabic literal amount recognition. Engineering Applications of Artificial Intelligence 19(1) (2006)Google Scholar
- 13.Singh, F., Singh, M.A.: dynamic classifier selection and combination approach to image region labelling, Signal Process. In: Image Commun., vol. 20, pp. 219–231 (2005)Google Scholar
- 14.Woloszynski, T., Kurzynski, M.: A Measure of Competence Based on Randomized Reference Classifier for Dynamic Ensemble Selection. In: ICPR 2010, Turkey, August 23-26, pp. 4194–4198 (2010)Google Scholar
- 15.Pechwitz, M., Maergner, V.: Baseline estimation for arabic handwritten words. In: Frontiers in Handwriting Recognition, pp. 479–484 (2002)Google Scholar