Title :
Performance optimization of neural networks in handwritten digit recognition using Intelligent Fuzzy C-Means clustering
Author :
Miri, E. ; Razavi, Seyed Mohsen ; Sadri, J.
Author_Institution :
Dept. of Electr. Eng., Univ. of Birjand, Birjand, Iran
Abstract :
In this paper, a new approach has been proposed in order to optimize performance of Multi Layer Perceptron Neural Networks in handwritten digit recognition. In the proposed approach, Fuzzy C-Means clustering with PSO optimizer has been used, and it has been applied in handwritten Farsi digits recognition. Obtained results show that with the help of this approach we can reduce the rate of misclassifications as compared to other common approaches found in the literature.
Keywords :
fuzzy set theory; handwritten character recognition; multilayer perceptrons; natural language processing; optimisation; pattern clustering; performance evaluation; PSO optimizer; handwritten Farsi digit recognition; intelligent fuzzy C-mean clustering; multilayer perceptron neural networks; neural networks; performance optimization; Feature extraction; Handwriting recognition; Neural networks; Neurons; Training; Vectors; Fuzzy C-Means (FCM) Clustering; Multi Layer Perceptron (MLP) Neural Network; OCR; PSO; Pattern Recognition; Recognition of Farsi Handwritten Digits;
Conference_Titel :
Computer and Knowledge Engineering (ICCKE), 2011 1st International eConference on
Conference_Location :
Mashhad
Print_ISBN :
978-1-4673-5712-8
DOI :
10.1109/ICCKE.2011.6413342
