Persian handwritten classification base on ensemble learning classifiers SVM،MLP
Persian handwritten classification base on ensemble learning classifiers SVM،MLP
Somaye Esmaeili Rad1 Amir Rajabi Behjat2 Mohammad Ali Bagheri3
1) Iran islamic Azad University of Rafsanjan Email:
2) Iran assistant professor in Islamic Azad university
3) Iran islamic Azad University of Rafsanjan
IOI Code
:
XYHD-BFZBB
Publication :
International Conference On Research Science And Technology(restconf.org)
Abstract :
Text classification is one of machine learning techniques, those based on content. The classifier is usually made in a learning process. Due to significant increase in the volume of text content, Automatic tools and methods have been noticed very much for document management and text content.
One of these tools is text classification, Which can be very effective in the field. Text classification has many applications in the systems of literature review; such as the search engines, filtering systems, spam detection, digital libraries and other similar systems. In this research, the tool of text classification was used on the configure of Hamshahri to assess Persian texts and texts were classified by combining SVM and KNN algorithms.we have upgrade this classifier by 540 Persian texts which had been taken from the configure of Hamshahri and examined on 70 similar texts. We observed that texts were divided into 7 categories . Among these categories, text of the sport, medical and political category were classified Better than others .The algorithms we used to measure accuracy of combining classifiers in the best case was reached to 0.98 for political context , Along with experimental results on the configure of Hamshahri by use of SVM and KNN algorithms indicated that the combination of algorithms and feature selection methods, while reduces the number of features, improves the efficiency and accuracy in the combining classifiers system.
Keywords : text classification algorithm knn algorithm svm