International Journal of Pattern Recognition and Artificial Intelligence

In order to give attention to the smooth distortion of two image mosaic evaluation indices concurrently, an image multiscale mosaic method based on fuzzy integral is proposed in this paper. Aiming at problems resulting from evaluating mosaic effect of single-factor evaluation index, the integrative evaluation method of the mosaic image based on synthesizing two independent single-factor indices is put forward by using a fuzzy integral which is largely subject to human feeling. According to the optimal image mosaic rules, an optimal weighted function is obtained. With the help of the optimal weighted function, wavelet coefficients are fused at different resolution levels. As a result, a smooth and seamless mosaic image is reconstructed. Finally, experiments are performed on the method presented in this paper.

By using the restricted and complementary relationship of the principle and secondary indexes, providing the description of the compound quantification of the fuzzy number, and analyzing the essential characteristic of fuzzy decision, we propose a kind of fuzzy genetic algorithm based on the principle index (PO-FGA for short) to deal with the fuzzy optimization and programming problems with fuzzy coefficients, fuzzy variables and fuzzy constraints. The concrete solution method is presented in accordance with the strategy of the unconditional penality transformation with conditional constrains. Then consider its convergence by using Markov chain theory and analyze its performance through two examples. All these indicate that this kind of algorithm is of faster speed of convergence, smaller number of iterations, has lower chances of trapping into the state of premature convergence and can be widely used in many problems of optimization.

The smart card is being used all over the world, because of its power to store data securely and execute calculations confidentially. This paper proposes an easy way to implement the strategy which can guarantee data integrity and privacy in the transmission path between the terminal application of the smart card and the smart card. This strategy consists of cryptographic functions of message and methods for transmitting APDU and managing secret keys. Some cryptographic functions of message are defined and developed for transmitting APDU confidentially. These cryptographic functions can keep the byte length of input and output messages identical. This strategy for transmitting APDU points out how to call cryptographic functions to encode or decode the data body of APDU, and how to construct a modified APDU message structure according to the original APDU message structure, which will be sent to its intended recipient instead of the original APDU message structure. With the proper secret keys, the authorized party can recover this modified APDU to its corresponding original APDU properly, without any information leakage. The transfer of secret keys to other parties is the most difficult aspect of secure APDU transmission, whereas data encryption/decryption is relatively straightforward. Two distinct strategies for managing secret keys, namely the static and the dynamic are introduced and compared. The dynamic strategy is preferred for a terminal application and smart card which do not need to authenticate each other. The implementation of this dynamic strategy has been detailed.

The detection of insulators contamination is difficult in power systems because many factors can influence the pollution. The contamination condition of insulators is usually estimated by detecting the root mean square (r.m.s) of surface leakage current via online-monitoring system. It ignores the influence of environmental factors, such as temperature, humidity, etc. As these factors are fuzzy-characterized, a new method based on Fuzzy Neural Network (FNN) is proposed to improve traditional insulation contamination detection. The renewed structure of FNN is put forward. The evaluation of contamination severity of insulators is achieved through FNN, which are trained by the field samples. The results prove the validity of the method proposed in the paper and can be used to eliminate the insulator from flashover fault and improve the condition-based maintenance (CBM).

Beam search algorithm, as an adaptation of branch and bound method, is regarded as one of the effective approaches in solving combinational optimization problems. In this paper, a new beam search algorithm for the large-scale permutation flow shop scheduling problem (FSP) is proposed. A new branching scheme is addressed and compared with the traditional branching scheme. With the new branching scheme, the number of partial schedules in the search tree can be greatly reduced. Based on a simple simulated annealing algorithm, partial schedules are globally evaluated. Numerical experiments show that good solutions of large-scale FSPs could be found with the proposed algorithm in a short time.

Formula extraction is the first stage of optical formula recognition which converts printed scientific documents into their corresponding electronic format. So far, little research has been done in this area. In this paper, an improved method using fuzzy classification and irregularity rate feature is proposed to separate formulas from texts in the printed documents. Firstly, according to a statistical threshold of distance, connected components are extracted and merged to form the areas of characters and lines. Secondly, the isolated formulas are extracted based on the line features. Finally, the formula symbols in the rest lines are labeled using irregularity degree, and the embedded formulas are located by extending kernel symbols using the propagation of context. In these steps, fuzzy classification algorithm and irregularity degree feature are introduced to solve the problems existing in traditional methods and improve the extracting accuracy. The experimental results show that the method is of great significance in both theory and practice.

Hyper Surface Classification (HSC), which is based on Jordan Curve Theorem in Topology, has proven to be a simple and effective method for classifying a larger database in our previous work. To select a representative subset from the original sample set, the Minimal Consistent Subset (MCS) of HSC is studied in this paper. For HSC method, one of the most important features of MCS is that it has the same classification model as the entire sample dataset, and can totally reflect its classification ability. From this point of view, MCS is the best way of sampling from the original dataset for HSC. Furthermore, because of the minimum property of MCS, every single deletion or multiple deletions from it will lead to a reduction in generalization ability, which can be exactly predicted by the proposed formula in this paper.

This paper proposes a fast fuzzy classifier of multicategory support vector machines (FMSVM) based on support vector domain description (SVDD). The main idea is that the proposed FMSVM is obtained by directly considering all data in one optimization formulation, using a fuzzy membership to each input point. The fuzzy membership is determined by support vector domain description (SVDD). For making support vector machine (SVM) more practical, we use an implement of the modified sequential minimal optimization (SMO) that can quickly solve SVM quadratic programming (QP) problems without any extra matrix storage or the use of numerical QP optimization steps at all. Compared with the existing SVMs, the newly proposed FMSVM that uses the L₂-norm in the objective function shows improvement with regards to accuracy of classification and reduction of the effects of noises and outliers. The experiment also shows the efficiency of the modified SMO for expediting the training of SVM.

We had developed the localized generalization error model for supervised learning with minimization of Mean Square Error. In this work, we extend the error model to Single Layer Perceptron Neural Network (SLPNN) and Support Vector Machine (SVM) with sigmoid kernel function. For a trained SLPNN or SVM and a given training dataset, the proposed error model bounds above the error for unseen samples which are similar to the training samples. As the major component of the localized generalization error model, the stochastic sensitivity measure formula for perceptron neural network derived in this work has relaxed the assumptions of same distribution for all inputs and each sample perturbed only once in previous works. These make the sensitivity measure applicable to pattern classification problems. The stochastic sensitivity measure of SVM with Sigmoid kernel is also derived in this work as a component of the localized generalization error model. At the end of this paper, we discuss the advantages of the proposed error bound over existing error bound.

Multiple Classifier System (MCS) is a very popular research topic in recent years. It has been proved theoretically and empirically to be better than single classifiers in many scenarios. Creating diverse sets of classifier is one of the key issues in building MCSs. Feature grouping is one of the methods to create diverse classifiers and it has been shown to improve the accuracy of an MCS. In this paper, we propose a new feature grouping method based on Genetic Algorithm (GA) with the localized Generalization Error Model as the evaluation criterion. The combined individual classifiers using the weighted sum are examined in this paper. Moreover, several feature grouping methods are compared with the proposed method in this work. The experimental results on benchmark dataset show that the MCS trained by the proposed method is promising.

To evaluate the pronunciation skills of spoken English is one of the key tasks for computer-aided spoken language learning (CALL). While most of the researchers focus on improving the speech recognition techniques to build a reliable evaluation system, another important aspect of this task has been ignored, i.e. the pronunciation evaluation model that integrates both the reliabilities of existing speech processing systems and the learner's pronunciation personalities. To take this aspect into consideration, a Sugeno integral-based evaluation model is introduced in this paper. At first, the English phonemes that are hard to be distinguished (HDP) for Chinese language learners are grouped into different HDP sets. Then, the system reliabilities for distinguishing the phonemes within a HDP set are computed from the standard speech corpus and are integrated with the phoneme recognition results under the Sugeno integral framework. The fuzzy measures are given for each subset of speech segments that contains n occurrences of phonemes within a HDP set. Rather than providing a quantity of scores, the linguistic descriptions of evaluation results are given by the model, which is more helpful for the users to improve their spoken language skills. To get a better performance, generic algorithm (GA)-based parameter optimization is also applied to optimize the model parameters. Experiments are conducted on the Sphinx-4 speech recognition platform. They show that, with 84.7% of average recognition rate of the SR system on standard speech corpus, our pronunciation evaluation model has got reasonable and reliable results for three kinds of test corpora.

This paper gives an initial study on the comparison between Genetic Algorithm (GA) and Simulated Annealing Algorithm (SAA). Firstly, a new algorithm is presented. This method combines Genetic Algorithm and Simulated Annealing Algorithm, and it can be used to optimize the three parameters α, β and γ. It involes the rules that are extracted from Fuzzy Extension Matrix (FEM). These parameters play an important part in the entire process of rule extraction based on FEM. Secondly, it provides some theoretical support to the direct selection of the parameter values through experiments. Lastly, five data sets from the UCI Machine Learning centers are employed in the study. Experimental results and discussions are given.

Considering the decreasing performance of robotic systems under the constraints of actuators, this paper concludes the rules of dynamic control process and extracts the knowledge of optimizing the output of robotic controller based on the analysis of a Lyapunov function. Then fuzzy rules are used to express the knowledge and embedded in the controller to direct the dynamic control process. Under this controller, the systemic requirement of high-level outputs of the actuators is limited to a great extent, at the same time the convergent performance is optimized by a fuzzy-sets method from which the saturated fault tolerant control is realized in robotic manipulators. Simulating results on a 2-DOF robot validate the effectiveness of the given controller.

Based on some results of the fuzzy network computation and competitive analysis, the Online Fuzzy Most Reliable Path Problem (OFRP), which is one of the most important problems in network optimization with uncertainty, has been originally proposed by our team. In this paper, the preliminaries about fuzzy and the most reliable path and competitive analysis are given first. Following that, the mathematical model of OFRP, in which two kinds of uncertainties, namely online and fuzzy, are combined to be considered at the same time, is established. Then some online fuzzy algorithms are developed to address the OFRP and the rigorous proofs of the competitive analysis are given in detail. Finally, some possible research directions of the OFRP are discussed and the conclusions are drawn.