Research on neural network has witnessed impressive progress on its model, learning algorithm and application. Owe to the ambiguity, fuzziness and ability of approximating functions of any complexity it provides, neur...
详细信息
ISBN:
(纸本)0780378652
Research on neural network has witnessed impressive progress on its model, learning algorithm and application. Owe to the ambiguity, fuzziness and ability of approximating functions of any complexity it provides, neural computing method has gained great success in various fields. However, serious limitations remain accompanying the method concerning accessibility, flexibility, scaling and reliability. Efforts have been made to relieve the problem, of which decomposition approach is typically used. Although it works, the method is greatly confined by the decomposition strategy and scalability of such system which is scarified. Previous researches focus on effectively dividing the input space to build up modular systems for given tasks and that the design of the modules themselves is neglected. In this paper, we introduce the notion of autonomous artificial neural network (AANN) for the first time and use AANN units as the basic building block of a modular neural system. The difference between AANN units and conventional neural network is that AANN units possess the ability of self-reflection. Unlike conventional neural networks, which directly give an output corresponding to an input without commenting on its quality, an AANN unit comments in its own point of view on the output, providing extra information such as whether the result is correct or not or to what degree the result can be believed. With AANN units as building blocks, a modular neural system is greatly relieved from what it suffers. Scalability, reliability and flexibility of such a system are greatly improved. We have also proposed a coding method to implement AANN units. And a neural network ensemble of AANN units is built up, which acquires knowledge progressively by inheriting the learned knowledge of AANN units attached to it.
Load balancing strategy has been founded and applied widely as an effective measure of performance optimization for most conventional distributed systems. However, for the sake of distributed embedded system's inh...
详细信息
ISBN:
(纸本)0780378652
Load balancing strategy has been founded and applied widely as an effective measure of performance optimization for most conventional distributed systems. However, for the sake of distributed embedded system's inherent properties, especially for real-time property, load balancing strategy may not be desirable. On consideration of those properties, an improved load management scheme is presented based on a new way of resource reservation, which has considerable enhancements in satisfying real-time requirements. First, the conception of satisfaction to real-time requirement of incoming task is introduced in a new aspect. Then the implementation of resource reservation is narrated in detail. Under this concept of load management, the new incoming tasks are allocated to the most heavily loaded node, if only the service node can provide real-time satisfaction. Thus, a great amount of calculating resources is reserved for those tasks that require almost full resource of a service node. In addition, a central control model based on this improved load management is introduced into application and implement in building up a prototype system according to this model. During this simulation, the new load management strategy based on resource reservation shows great advantage according to real-time criteria against load balancing. The conclusion comes into being that the resource reservation strategy arising in the paper benefits from a distributed embedded system with high real-time requirements.
In the paper proposes a granular logic, which defines the syntax and semantics of the logic. The formula in the logic is a granule or any Boolean combination of the granules. This paper discusses a reasoning based on ...
详细信息
ISBN:
(纸本)0780375084
In the paper proposes a granular logic, which defines the syntax and semantics of the logic. The formula in the logic is a granule or any Boolean combination of the granules. This paper discusses a reasoning based on the granular computing. The reasoning is used in expert systems, and is also an extension of match reasoning in production rules. The inclusion and closeness degree between granules are defined in the paper. Finding the match among granules satisfying given closeness degree finishes the approximate reasoning. The paper defines the granular operation that is both logic and set theory. Finally, the matching reasoning and deductive inference is illustrated with real examples.
The hyperlink structure is an important resource in web information retrieval, especially for the ranking algorithm in search engines. This paper investigates the current main researches for the hyperlink structure in...
详细信息
ISBN:
(纸本)0780375084
The hyperlink structure is an important resource in web information retrieval, especially for the ranking algorithm in search engines. This paper investigates the current main researches for the hyperlink structure in web IR;proposes our own ideas for the hyperlink analysis to expand the HITS algorithm, add the webpage quality factors into the hyperlink, gives the expanding algorithm and proves the efficiency and convergence of our algorithm.
In this paper, a new equation of saturation in RGB-to-HSI conversion is proposed on the basis of HSI Color Space, aimed to provide more rapidity of computing in real-time control systems due to fewer operations needed...
详细信息
ISBN:
(纸本)0780375084
In this paper, a new equation of saturation in RGB-to-HSI conversion is proposed on the basis of HSI Color Space, aimed to provide more rapidity of computing in real-time control systems due to fewer operations needed. The theatrical validness of new conversion equation is studied and testified. Besides its significant advantages over traditional equation on temporal performance, better mapping performances is mathematically explained as a minor advantage. It is shown in illustrative experiments that average 30% time was saved in computing when using new saturation conversion equation instead of traditional one.
Automated acquisition of rules is a very important issue in knowledge engineering. In this paper, a new method for generating fuzzy rules from numerical data for fuzzy classification problems based on the correlation ...
详细信息
ISBN:
(纸本)0780375084
Automated acquisition of rules is a very important issue in knowledge engineering. In this paper, a new method for generating fuzzy rules from numerical data for fuzzy classification problems based on the correlation analysis between two fuzzy sets has been proposed. Experimental results and comparison with Yuan and Shaw's method are presented as well.
Iris verification is one of the biometrics verification technologies. This paper proposes a new iris verification method based on fractional Fourier transform. Through comparing two irises' Fractional Fourier Tran...
详细信息
ISBN:
(纸本)0780375084
Iris verification is one of the biometrics verification technologies. This paper proposes a new iris verification method based on fractional Fourier transform. Through comparing two irises' Fractional Fourier Transform, we can distinguish the people whether they are the same person. At last, we introduce some applications of iris verification used for security in E-commerce.
Owing to the computational complexity requirement, pruning a fully connected recurrent neural network (RNN) would be ineffective for large size RNN. In paper [13], several non-heuristic pruning algorithms for fully co...
详细信息
ISBN:
(纸本)0780375084
Owing to the computational complexity requirement, pruning a fully connected recurrent neural network (RNN) would be ineffective for large size RNN. In paper [13], several non-heuristic pruning algorithms for fully connected RNN have been proposed, some of them are extended from EKF based approaches and some of them are based on weight magnitude, together with some tricks on the pruning procedures. Their effectiveness on their computational complexities have been analyzed. In this paper, their effectiveness on network sizes and the generalization abilities, will be evaluated.
The dither matrix has been extensively applied in image processing, color matching and other aspects. In this paper, a new concept is put forward about generalized Latin square and a mathematical description about the...
详细信息
ISBN:
(纸本)0780375084
The dither matrix has been extensively applied in image processing, color matching and other aspects. In this paper, a new concept is put forward about generalized Latin square and a mathematical description about the dither matrix, the counting formula of generalized Latin square and a pair of orthogonal generalized Latin squares are given. The necessary and sufficient conditions of two orthogonal generalized Latin squares are presented and one-to-one correspondence relation of two sets between dither matrix and a pair of orthogonal generalized Latin squares are established by studying the generalized Latin squares. So the study of dither matrix can be transformed to that of a pair of orthogonal generalized Latin squares.
With the increasing worldwide availability of the Internet and the web content being accessible internationally and mobility, the e-learning industry faced a tremendous development from off-line to on-line in recent y...
详细信息
ISBN:
(纸本)0780375084
With the increasing worldwide availability of the Internet and the web content being accessible internationally and mobility, the e-learning industry faced a tremendous development from off-line to on-line in recent years and the model is expanded to support m-learning for thin-clients devices. However, CJK data handling in knowledge management is still a critical issue for m-learning development because most thin-client technology is developed in Western countries that support only alphabet-based languages, while the number of CJK users increases greatly. In this paper, a novel technique is presented to enhance the interaction and communication between the teachers and students in the m-learning platform by using multi-agent. The Student Agents are responsible to organize and analyze the learning activities, then carry these CJK knowledge flows and report to the Instructor Agents. By using these globalized user client agents to monitor the study progress of the students in the multi-agent platform, it not only allows instructors to identify the actual need for each student beside of the student's direct request, but also help the developer to delivery international m-learning application. Therefore, the me-learning environment becomes truly dynamic and globalized.
暂无评论