Recognition of handwritten characters using modified fuzzy hyperline segment neural network

Review on Classification and Clustering using Fuzzy Neural Networks

“…The contributors of [14] have utilized the modified FHLSNN for recognition of handwritten characters. The modified FHLSNN is a variant with few minor modifications in the learning algorithm of FHLSNN.…”

“…Modified fuzzy hyperline segment neural network [14] supervised learning, uses fuzzy set hyperline segments, online training, modified membership function, quick learning, exceptionally quick in recall, provides soft decision Modular Fuzzy Hypersphere Neural Network [15] offers higher degree of parallelism, each module exposed to the patterns of only one class, extremely quick in training, suitable for big real database, online training A fuzzy min-max neural network classifier with compensatory neuron architecture [18] uses hyperbox fuzzy sets, employs new compensatory neuron architecture, supports on line adaptation in a single pass, yields reduced classification and gradation errors, performance is less dependent on the initialization of maximum hyperbox size coefficient A modified fuzzy min-max neural network with rule extraction [20] uses an Euclidean distance measure for prediction, a rule extraction algorithm, uses pruning A modified fuzzy min-max neural network with A genetic-algorithmbased rule extractor for pattern classification [22] a two-stage pattern classification and rule extraction system, a modified FMM neural network, utilizes genetic-algorithm (GA)-based rule extractor Data-core-based fuzzy min-max neural network [23] a new membership function, contraction process is eliminated A hybrid FMM-CART model [24] utilizes FMM for classification and CART for rule extraction, supports offline and online properties An enhanced fuzzy min-max neural Network for pattern classification [25] a new hyperbox expansion rule, extended hyperbox overlap test rule, a new hyperbox contraction rule…”

“…These advantages of FNNs with their proven record has caused a great amount of interest and their use in pattern classification and recognition problems [4], [6], [7], [8], [11], [14], [15], [16], [17], [18], [20], [21], [22], [23], [24], [25], [26].…”

Fourier fuzzy neural network for clustering of objects based on the gross shape and its application to handwritten character recognition

“…The membership function designed in (6) possesses all the properties of fuzzy sets like normality and convexity described in [10]. The plot of membership function for a 2-D cluster prototype (0.5, 0.5), with y = I and y = 4 are shown in Fig.…”

“…Quad vectors used for feature extraction is found superior than the ring features. Kulkarni et al [5] proposed Fuzzy hyperline segment neural network (FHLSNN) for rotation invariant handwritten character recognition using ring features [2] and magnitude of Zernike moments proposed by Khotanzad [7,8]. Fuzzy Neural Network (FNN) proposed by Kawn and Cai [9] has capability to extract image features as well as to recognize them.…”