Research on Bayesian Network Based User＇s Interest Model

It has very realistic significance for improving the quality of users＇ accessing information to filter and selectively retrieve the large number of information on the Internet. On the basis of analyzing the existing users＇ interest models and some basic questions of users＇ interest （representation, derivation and identification of users＇ interest）, a Bayesian network based users＇ interest model is given. In this model, the users＇ interest reduction algorithm based on Markov Blanket model is used to reduce the interest noise, and then users＇ interested and not interested documents are used to train the Bayesian network. Compared to the simple model, this model has the following advantages like small space requirements, simple reasoning method and high recognition rate. The experiment result shows this model can more appropriately reflect the user＇s interest, and has higher performance and good usability.     

An Intrusion Alarming System Based on Self-Similarity of Network Traffic

Intrusion detection system can make effective alarm for illegality of network users, which is absolutely necessarily and important to build security environment of communication base service. According to the principle that the number of network traffic can affect the degree of self-similar traffic, the paper investigates the variety of self-similarity resulted from unconventional network traffic. A network traffic model based on normal behaviors of user is proposed and the Hurst parameter of this model can be calculated. By comparing the Hurst parameter of normal traffic and the self-similar parameter, we can judge whether the network is normal or not and alarm in time.     

Study of a new method for three-dimensional restoration from single image

The aim of threedimensional recovery technology from the image is to recover the relative height of each point on the surface from the light variations in the single image and carry out the recovery. A new method that recovers threedimensional is presented object based on radius basis function for the image from the unknown light source direction, which constructs a surface equation by the network, uses the reflectivity function as a constraint, continuonsly estimates the light source direction in the selflearning process of the network from the bright spots around the spread of the image, and eventually obtains a satisfactory surface equation. This method makes the resumption of surface good continuity and smoothness, and can recover the height value of each network point of the image and be automatically inserted in any point the among the network. It is suitable for the image of Lambert reflection model and the image of the Specular reflection model and the mixedreflex model.     

Non-cooperative game for network access selection in heterogeneous integrated networks

The integration of different heterogeneous access networks is one of the remarkable characteristics of the next generation network,in which users with multi-network interface terminals can independently select access network to obtain the most desired service.A kind of unified quantification model of non-monotone quality of service （QoS） and a model of non-cooperative game between users and networks are proposed for heterogeneous network access selection.An optimal network pricing mechanism could be formulated by using a novel strategy which is used in this non-cooperative game model to balance the interests of both the users and the networks.This access network selection mechanism could select the most suitable network for users,and it also could provide the basis when formulating QoS standards in heterogeneous integrated networks.The simulation results show that this network selection decision-making algorithm can meet the users＇ demand for different levels service in different scenes and it can also avoid network congestion caused by unbalanced load.     

Fuzzy ARTMAP neural network for seafloor classification from multibeam sonar data

This paper presents a seafloor classification method of multibeam sonar data, based on the use of Adaptive Resonance Theory （ART） neural networks. A general ART-based neural network, Fuzzy ARTMAP, has been proposed for seafloor classification of multibeam sonar data. An evolutionary strategy was used to generate new training samples near the cluster boundaries of the neural network, therefore the weights can be revised and refined by supervised learning. The proposed method resolves the training problem for Fuzzy ARTMAP neural networks, which are applied to seafloor classification of multibeam sonar data when there are less than adequate ground-troth samples. The results were synthetically analyzed in comparison with the standard Fuzzy ARTMAP network and a conventional Bayesian classifier. The conclusion can be drawn that Fuzzy ARTMAP neural networks combining with GA algorithms can be alternative powerful tools for seafloor classification of multibeam sonar data.     

Experimental study of fatigue degree quantification for multi-feature fusion identification

A comprehensive quantification method of fatigue degree is proposed concerning subjective and objective quantifications. Using the fatigue degree test software, fatigue degree is objectively quanti- fied by analyzing the reaction and operation abilities of drivers about traffic signals. By comparison experiment with that EEG signal based, multivariate statistical analysis and fusion identification based on BP neural network （ BPNN） results show that the experimental procedure is simple and practical, and the proposed method can reveal the correlation between fatigue feature parameters and fatigue degree in theory, and also can achieve accurate and reliable quantification of fatigue degree, especially under the associated action of multiple fatigue feature parameters.     

A New Approach for Lossless Image Compression Based on Fuzzy Adaptive Prediction

This paper proposes a novel approach for image lossless compression based on fuzzy logic and adap-tive prediction. By a flexible strategy, the method can acquire a set of original predictors describing the more detail characteristic. Using a neural network, the proposed method can more efficiently organize the training of original predictors and implement adaptive prediction in fuzzy style. In entropy coding phase,the context-based conditional adaptive arithmetic encoding is adopted. The experiments demonstrate the characteristics make the approach achieve good tradeoff between computational complexity and efficiency of prediction and good performance for lossless compression.     

A method for measuring the clock offset of two hosts in the network

In order to detect the performance parameters of the network, for example, the network delay or delay jitter, the clock synchronization relations between the two hosts at two ends along the network must be calculated in advance. Then with the correct temporal relations between the two hosts, multimedia transmission along the network and display can occur by the proper order. A refined method based on Paxson's algorithm is proposed and testified. More accurate results can be attained by the method. By the way, the method can be used in a more complicated environment. Furthermore, an end-to-end network performance tester based on the proposed algorithm is designed and implemented.     

Improved Genetic Algorithm Application in Textile Defect Detection

Based on an efficient improved genetic algorithm, a pattern recognition approach is represented for textile defects inspection. An image process is developed to automatically detect the drawbacks on textile caused by three circumstances： break, dual, and jump of yams. By statistic method, some texture feature values of the image with defects points can be achieved. Therefore, the textile defects are classified properly. The advanced process of the defect image is done. Image segmentation is realized by an improved genetic algorithm to detect the defects. This method can be used to automatically classify and detect textile defects. According to different users＇ requirements, different types of textile material can be detected.     

ACTPred： Activity Prediction in Mobile Social Networks

A current trend for online social networks is to turn mobile.Mobile social networks directly reflect our real social life,and therefore are an important source to analyze and understand the underlying dynamics of human behaviors （activities）.In this paper,we study the problem of activity prediction in mobile social networks.We present a series of observations in two real mobile social networks and then propose a method,ACTPred,based on a dynamic factor-graph model for modeling and predicting users＇ activities.An approximate algorithm based on mean fields is presented to efficiently learn the proposed method.We deploy a real system to collect users＇ mobility behaviors and validate the proposed method on two collected mobile datasets.Experimental results show that the proposed ACTPred model can achieve better performance than baseline methods.     

Performance Predictable ServiceBSP Model for Grid Computing

This paper proposes a performance prediction model for grid computing model ServiceBSP to support developing high quality applications in grid environment. In ServiceBSP model, the agents carrying computing tasks are dispatched to the local domain of the selected computation services. By using the IP （integer program） approach, the Service Selection Agent selects the computation services with global optimized QoS （quality of service） consideration. The performance of a ServiceBSP application can be predicted according to the performance prediction model based on the QoS of the selected services. The performance prediction model can help users to analyze their applications and improve them by optimized the factors which affects the performance. The experiment shows that the Service Selection Agent can provide ServiceBSP users with satisfied QoS of applications.     

Anomaly-based model for detecting HTTP-tunnel traffic using network behavior analysis

Increasing time-spent online has amplified users＇ exposure to tile tilreat oI miormanon leakage. Although existing security systems （such as firewalls and intrusion detection systems） can satisfy most of the security requirements of network administrators, they are not suitable for detecting the activities of applying the HTTP-tunnel technique to steal users＇ private information. This paper focuses on a network behavior-based method to address the limitations of the existing protection systems. At first, it analyzes the normal network behavior pattern over HTI＇P traffic and select four features. Then, it pres- ents an anomaly-based detection model that applies a hierarchical clustering technique and a scoring mechanism. It also uses real-world data to validate that the selected features are useful. The experiments have demonstrated that the model could achieve over 93% hit-rate with only about 3% false- positive rate. It is regarded confidently that the approach is a complementary technique to the existing security systems.     

Recognition algorithm of seabed pipeline defect inspection based on dynamic WBF neural networks

This paper describes a magnetic flux leak (MFL) model of pipeline defect inspection, and presents a recognition algorithm based on dynmnic wavelet basis function (WBF) neural network. The dynamic network utilizes multiscale and multireselution orthogonal wavelet, through signals backwards propagation:,has more significant advantages than BP or other neural networks used in MFL inspection. It also can control the accuracy of the predicted defect profiles, high-speed convergence possessing and well approaching feature. The perfonnance applying the algorithm based on the network to predict defect profile from experimental MFL signals is presented.     

Theory,Method and Application of a State RecognitionBased on Theory of Evidence

A new method of state recognition based on the theory of evidence was proposed. By this method, the plausible function which the sample awaiting recognition belongs to each category can be obtained through distance function. For the marginal samples,two or a batch of evidences can be combined and a new plausible function can be obtained by new evidence. Then the categories of samples can be determined according to plausibility function. This method provides a beder reasoning framework. The result proves the rate of recoghition correctness.     

An E.Commerce Recommender System Based on Content-Based Filtering

Content-based filtering E-commerce recommender system was discussed fully in this paper. Users＇ unique features can be explored by means of vector space model firstly. Then based on the qualitative value of products informa tion, the recommender lists were obtained. Since the system can adapt to the users＇ feedback automatically, its performance were enhanced comprehensively. Finally the evaluation of the system and the experimental results were presented.     

Artificial neural network prediction of mechanical properties of hot rolled low carbon steel strip

Conventionally, direct tensile tests are employed to measure mechanical properties of industrially pro- duced products. In mass production, the cost of sampling and labor is high, which leads to an increase of total pro- duction cost and a decrease of production efficiency. The main purpose of this paper is to develop an intelligent pro- gram based on artificial neural network （ANN） to predict the mechanical properties of a commercial grade hot rolled low carbon steel strip, SPHC. A neural network model was developed by using 7 x 5 x 1 back-propagation （BP） neural network structure to determine the multiple relationships among chemical composition, product pro- cess and mechanical properties. Industrial on-line application of the model indicated that prediction results were in good agreement with measured values. It showed that 99.2 % of the products＇ tensile strength was accurately pre- dicted within an error margin of ~ 10 %, compared to measured values. Based on the model, the effects of chemical composition and hot rolling process on mechanical properties were derived and the relative importance of each in- put parameter was evaluated by sensitivity analysis. All the results demonstrate that the developed ANN models are capable of accurate predictions under real-time industrial conditions. The developed model can be used to sub- stitute mechanical property measurement and therefore reduce cost of production. It can also be used to control and optimize mechanical properties of the investigated steel.     

Approach to Weighted Geometric Evaluation Based on Projection Pursuit

Weighted geometric evaluation approach based on Projection pursuit （PP） model is presented in this paper to optimize the choice of schemes. By using PP model, the multi-dimension evaluation index values of schemes can be synthesized into projection value with one dimension. The scheme with a bigger projection value is much better, so the schemes sample can be an optimized choice according to the projection value of each scheme. The modeling of PP based on accelerating genetic algorithm can predigest the realized process of projection pursuit technique, can overcome the shortcomings of large computation amount and the difficulty of computer programming in traditional projection pursuit methods, and can give a new method for application of projection pursuit technique to optimize choice of schemes by using weighted geometric evaluation. The analysis of an applied sample shows that applying PP model driven directly by samples data to optimize choice of schemes is both simple and feasible, that its projection values are relatively decentralized and profit decision-making, that its applicability and maneuverability are high. It can avoid the shortcoming of subjective weighing method, and its results are scientific and objective.     

Simulation of Artificial Intelligence for Automotive Air-conditioning System

The artificial intelligence is applied to the simulation of the automotive air-conditioning system ( AACS) . According to the system's characteristics a model of AACS, based on neural network, is developed. Different control methods of AACS are discussed through simulation based on this model. The result shows that the neural-fuzzy control is the best one compared with the on-off control and conventional fuzzy control method. It can make the compartment's temperature descend rapidly to the designed temperature and the fluctuation is small.     

Evenness Control of Rapier Loom Tension by Using Neural Network

The neural network model （NNM） is used to control the evenness of rapier loom tension and to model the uncertain parameters of drafting process. This model can estimate recursively the tension and incorporate predictive control strategies. After controlling the effect testing, it can be seen that the behavior is excellent following different set point and variations, such as the diameter of the beam of weaver, air stream, density of warp and weft, type of textile, etc. An optimum steady state scheme has been fixed in keeping evenness of tension of loom. This method improves obviously the efficiency and quality of production.     

Edge detection of range images using genetic neural networks

Due to the complexity and asymmetrical illumination, the images of object are difficult to be effectively segmented by some routine method. In this paper, a kind of edge detection method based on image features and genetic algorithms neural network for range images was proposed. Fully considering the essential difference between an edge point and a noise point, some characteristic parameters were extracted from range maps as the input nodes of the network in the algorithm. Firstly, a genetic neural network was designed and implemented. The neural network is trained by genetic algorithm, and then genetic neural network algorithm is combined with the virtue of global optimization of genetic algorithm and the virtue of parallel computation of neural network, so that this algorithm is of good global property. The experimental results show that this method can get much faster and more accurate detection results than the classical differential algorithm, and has better antinoise performance.