A New Approach in Persian Handwritten Letters Recognition Using Error Correcting Output Coding

Classification Ensemble, which uses the weighed polling of outputs, is the art ofcombining a set of basic classifiers for generating high- performance, robust and morestable results. This study aims to improve the results of identifying the Persianhandwritten letters using Error Correcting Output Coding (ECOC) ensemble method.Furthermore, the feature selection is used to reduce the costs of errors in our proposedmethod. ECOC is a method for decomposing a multi-way classification problem intomany binary classification tasks; and then combining the results of the subtasks into ahypothesized solution to the original problem. Firstly, the image features are extractedby Principal Components Analysis (PCA). After that, ECOC is used for identificationthe Persian handwritten letters which it uses Support Vector Machine (SVM) as thebase classifier. The empirical results of applying this ensemble method using 10 real-world data sets of Persian handwritten letters indicate that this method has betterresults in identifying the Persian handwritten letters than other ensemble methods andalso single classifications. Moreover, by testing a number of different features, thispaper found that we can reduce the additional cost in feature selection stage by usingthis method.

کلیدواژه ها:

Persian Handwritten Letters Recognition ، Classification Ensemble ، Feature Selection