Applied Mechanics and Materials Vols. 475-476

Abstract: The research started from the adaboost algorithm, which was optimized from the training process and the detection process, and on this basis face detection and tracking system was developed.Adaboost is a main algorithm in using.It calculate easily and extraction rate rapidly.But it also have some problem to resolve,such as the detection time is too long,the samples number is so limited and so on.This article give some optimizition methods in the training procedure.

Abstract: Supplier selection is a multi-criteria decision making problem which includes evaluation factors. In order to select the best suppliers, it is crucial to considering the both qualitative and quantitative factors simultaneously. The analytic hierarchy Process is a kind of Decision-making method with the combination of quantity and quality. Combining with fuzzy algorithm, this paper established a supplier evaluating system based on ERP database, discussing many issues in the systematic design and developing the implement step. The system make the evaluation process transparent by using data mining and fuzzy-AHP algorithm, and avoid the disturbing from excessively subjective factors. Thus it can guarantee the fairness of the assessment. Finally, the example and conclusion are given.

Abstract: The P-sets consists of an internal P-set and an outer P-set . It has dynamic characteristics. In order to obtain a new method of generation and identification information, the paper proposes the concepts of-recursive genetic information and-recursive genetic degree by using the dynamic characteristic of the internal P-set. Also the structure and characteristics about-recursive genetic information are given, including the the identification theorems and the identification criterions for the-recursive genetic information. Finally, the application example of the-recursive genetic information is provided. The structure of the-recursive genetic information is a new mathematical tool to process informaiton that has inward contraction characteristic.

Abstract: In many practical applications, the impulse responses of the unknown system are sparse. However, the standard Least Mean Square (LMS) algorithm does not make full use of the sparsity, and the general sparse LMS algorithms increase steady-state error because of giving much large attraction to the small factor. In order to improve the performance of sparse system identification, we propose a new algorithm which introduces a variable step size method into the Reweighted Zero-Attracting LMS (RZALMS) algorithm. The improved algorithm, whose step size adjustment is controlled by the instantaneous error, is called Variable step size RZALMS (V-RZALMS). The variable step size leads to yielding smaller steady-state error on the premise of higher convergent speed. Moreover, the sparser the system is, the better the V-RZALMS performs. Three different experiments are implemented to validate the effectiveness of our new algorithm.

Abstract: The perspective-n-point (PnP) problem is originated from camera calibration. It is to determine the position and orientation of the camera with respect to a scene object from n correspondent points. And a new stable algorithm by using a geometric constraint called perspective similar triangle (PST) can give new equations to solve P3P. The PST method achieves high stability in the permutation problem and in presence of image noise. Using the complete discrimination system, we obtain the solution classification of the new equation for the P3P problem. The solution classification gives a set of formulas to determine the number of real solutions to the P3P problem. Based on the formulas, we may know whether the parameters give multiple solutions or not and are critical or not which is very important to present robust algorithm.

Abstract: Let π be a set of primes. Isaacs established the π-theory of characters, which generalizes the theory of Brauer module characters. Based on Isaacss work, we introduce the definition of Mπ-groups, and prove that if G=NwrC_p is an Mπ-group, where C_p is a cyclic group of order p and pπ, then N is an Mπ-group.

Abstract: In this paper, we present a new method for solving of the one dimensional Burgers equation, that is the space-time Chebyshev pseudospectral method. Firstly, we discretize the Burgers equation in one dimensional space with Chebyshev pseudospectral method. Finally, numerical results obtained by this way are compared with the exact solution to show the efficiency of the method. The numerical results demonstrate high accuracy and stability of this method.

Abstract: In this paper, we deal with the monotonicity properties of the first eigenvalue of the Laplacian operator. Firstly, by using the monotonicity formula of the F functional, we derive a monotonicity formula of the first eigenvalue of the Laplacian operator. Based on this, we also prove an exponential decreasing property of the first eigenvalue.

Abstract: Recommender systems are widely employed in many fields to recommend products, services and information to potential customers. As the most successful approach to recommender systems, collaborative filtering (CF) predicts user preferences in item selection based on the known user ratings of items. It can be divided into two main braches - the neighbourhood approach (NB) and latent factor models. Some of the most successful realizations of latent factor models are based on matrix factorization (MF). Accuracy is one of the most important measurement criteria for recommender systems. In this paper, to improve accuracy, we propose an improved MF model. In this model, we not only consider the latent factors describing the user and item, but also incorporate content information directly into MF.Experiments are performed on the Movielens dataset to compare the present approach with the other method. The experiment results indicate that the proposed approach can remarkably improve the recommendation quality.

Abstract: A new FB-function based on the P₀ function is given in this paper. The nonlinear complementarity problem is reformulated to solve equivalent equations based on the FB-function. A modified smooth Newton method is proposed for nonlinear complementarity problem. Under mild conditions, the global convergence of the algorithm is proved. The numerical experiment shows that the algorithm is potentially efficient.