Issue No. 07 - July (2008 vol. 30)
This study investigates Bayes classification of online Arabic characters represented by histograms of tangent differences and Gibbs modeling of the class-conditional probability density functions. The parameters of these Gibbs density functions are estimated following the Zhu, Wu, and Mumford constrained maximum entropy formalism, originally introduced for image and shape synthesis. We investigate two partition function estimation methods: one uses the training sample and the other draws from a reference distribution. The efficiency of the corresponding Bayes decision methods, and of a combination of these, is shown in experiments using a database of 9504 freely written samples by 22 scriptors. Comparisons to the nearest neighbor rule method and Kohonen neural network methods are provided.
Clustering, classification, and, association rules, Associative memories
Amar Mitiche, Neila Mezghani, Mohamed Cheriet, "Bayes Classification of Online Arabic Characters by Gibbs Modeling of Class Conditional Densities", IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 30, no. , pp. 1121-1131, July 2008, doi:10.1109/TPAMI.2007.70753