Jianghong Han

An improved adaptive Genetic Algorithm in Optimization of Partner Selection

Partner selection is a critical problem in organizing virtual enterprises according to 3 main indexes of cost, credit degree and make span provided by the candidates. An improved genetic algorithm (AGA) with total fitness ranking-based selection and adaptive operator is presented. Selection based on total fitness ranking make multi-objective problem several single- objective optimizations, insures rational interval between individuals and avoids premature convergence. Crossover and mutation operator are adjusted according to the fitness and iterative degree. Thus, each individual owns the ability of self-adaptation with the variation of environment. The simulated results verified the effectiveness of AGA.Recently, semantic smoothing is proposed as an efficient solution for the improvement of document cluster quality. However, the existing semantic smoothing model is not effective for partitional clustering to enhance the document clustering quality. In this paper, inspired by the TF*IDF schema and background elimination strategy, we first introduce an improved semantic smoothing model, which is suitable for both agglomerative and partitional clustering. Based on the improved semantic smoothing model, two model-document clustering algorithms, the partitional clustering algorithm and the agglomerative clustering algorithm, are also presented. The experimental results show our algorithms are more effective than the previous methods to improve the cluster quality.

Mining Consumers' Most Adaptive Products by Efficient Clustering Algorithm

Use clustering methods to discover the individual consumers most adaptive products, which can support to make better decisions of marketing service. First, oriented from the consumers transactional data that we will mine and targeted by finding some consumers most adaptive products, we present a simple and efficient cluster algorithm to put the most similar data into the same group. Then we can find the mined consumers most adaptive products from the cluster. Moreover, we propose a Boolean algorithm to improve the performance of the previous.

Research of Component-Based Hybrid Design Pattern for Real-Time Microkernel

Adopting the traditional object-oriented and component-based methods, the software architecture design pattern can be applied in the concurrent, realtime and distributed fields, which will replace the kind of strong-coupling and process-oriented development predominant in the real-time embedded software to satisfy the needs of the fast development of hardware and communication technology. Combined with the mechanistic design pattern put forward by Gamma et al. (like layer, microkernel, component, etc.), in the light of the view between the subsystem and the component architecture and of the concurrency, correct design pattern is generated by means of pattern incubating, mining or instantiation, which is then applied to develop a robust and scalable architecture for real-time system. As the Component-based Software Engineering (CBSE) can effectively enhance the quality and efficiency of the software development process, the technology is introduced to architecture design, offering mature applications by means of framework instantiation. Based on the hybrid design pattern, the architecture and the process of the software become mutually reinforcing and complementary, and both the interior quality and the process quality of the software are guaranteed.

Data Synchronization Infrastructure with Middleware

Networked applications with non-HTTP protocol are not universally supported by CLDC/MIDP. Infrastructure deployed with middleware services these MIDlets via HTTP protocol, implements isomerous data consistency between mobile and Web database, and the security is achieved by basic and digest access authentication in HTTP authenticate response. Finally, a prototype is established verified the feasibility of framework, and the security is achieved concretely through base64 and md5.

Clustering problem using adaptive genetic algorithm

Clustering is very important to data analysis and data minig. The K-Means algorithm, one of the partitional clustering approaches, is an iterative clustering technique that has been applied to many practical clustering problems successfully. However, the K-Means algorithm suffers from several drawbacks. In this paper, an adaptive genetic algorithm be present , it solve disadvantages of K-Means by combine parallel genetic algorithm, evolving flow and adaptive. Experimental results show that the adaptive genetic algorithm have advantages over traditional Clustering algorithm.

A facial expression classification algorithm based on principle component analysis

In this paper, we try to develop an analytical framework for classifying human basic emotions. We try to find out what are the major components of each facial expression, what are the patterns that distinguish them from one another. We applied widely used pattern recognition technique-principle component analysis to characterize the feature point displacements of each basic human facial expression for each individual in the existing database. For faces not existent in the database, so called “novel face” in our experiment, we will first find the face in the database that has most likely neutral face to this individual, and base on an assumption that are widely accepted in cognitive science, we will classify this novel face to the category where the most similar one belongs, and classifying his/her facial expression using the so called “expression model” of the most similar individual. This kind of approach has never be exploited before, then we will examine its robustness in our experiment.

Design of the multimedia communication protocol and system for wireless LAN

The protocol of wireless LAN —IEEE 802.11 put forward DCF/PCF to offer the real time transmission, but it did not meet the need of multimedia transmission, so a new MAC protocol is developed to solve this problem. In this paper we discuss the way of implementing and processing the multimedia data streams in WLAN. We regard the transmission of multimedia as multi-channel accessing with synchronization to study it. We start from the parameter QoS and mapped the multimedia transmission data streams to the different priorities, and use negotiation approach to allocate the resource.