Advanced Materials Research Vols. 989-994

Abstract: At present, the competition is increasingly fierce between the securities company, whether can effectively prevent the loss of users, reducing loss rate is a difficult problem at present each securities company urgently needs to solve. The model based on the principle of data mining, proposes a prediction method based on Logistic regression algorithm. Prediction model is built based on Logistic regression algorithm and the validity and accuracy of the model is verified by experiment, provides a new method and thinking for the securities company customer churn prediction.

Abstract: Pull and drag transportation is an advanced transportation organization ,it is lower transport costs, higher efficiency, better energy saving. Customer operation center is important nodes in the pull and drag transportation routes, related to the overall layout of the transportation network and the overall effectiveness. Using AHP in this article, take full account to needs of pull and drag transportation , and analysis the main influencing factors about customer operation center location from the perspective of society and the business. Established model, obtains the importance weight of the influence factors, and provide the basis for evaluation. This method makes the selection process efficient, reasonable, and has a certain practicality.

Abstract: Applying transportation planning theory and operations research method , this paper proposes the layout model of integrated freight hub with two-stage , which first builds a model to predict the attributes of integrated freight hub aiming for the optimization of regional freight transportation organization , and then works out the way to classify district integrated freight hub based on Hierarchical cluster. Eventually, take a specific example for empirical study.

Abstract: Distribution is an important function of enterprise logistics. Through efficient distribution activities, it can reduce the total cost of logistics enterprises and improve logistics efficiency and service levels. Therefore, the rationality of the logistics and distribution system design will directly affect the survival and development of enterprises. In this dissertation uses a retail business distribution network for the study, to study its distribution optimization models and methods, using a variety of algorithms for analysis, to understand its distribution network efficient mode of operation.

Abstract: The main idea of rough set theory is to extract decision rules by attribute reduction and value reduction in the premises of keeping the ability of classification. This paper presents the design of model for customer division based on rough set, and uses algorithms for attribute reduction and rule extraction in rough set to analyze the customer of supermarket. This paper also introduces how to achieve the minimum result of attribute reduction and decision-making via decision-making report, winkling redundant attribute and over-rule of decision.

Abstract: Event-driven investments have gained great importance and popularity. Due to the importance of the timely and effective messages for successful investment, the automated categorization of documents into predefined labels has received an ever-increased attention in the recent years. This paper implements a new text document classifier by integrating the K-nearest neighbour (KNN) classification approach with the VSM vector space model. By screening the feature items and weighted key items, the proposed classifier turns the financial information text into N-dimensional vector and identified the positive and negative information, furthermore achieve to the classification optimized. In addition, the classification model constructed by the proposed algorithm can be updated incrementally, and it has great scalability in event-driven securities investment for investors.

Abstract: Intelligent warehouse management can effectively improve the work efficiency and decreasing operating costs for a company, and the reasonableness of warehouse layout has played a very important role in it. This paper will aim at the warehouse layout problem, analyze the superiority of genetic algorithm and dynamic programming for the problem, and base on the genetic algorithm, integrating the thoughts of dynamic programming to put forward an optimization strategy. It can improve the accuracy and efficiency of genetic algorithm by a modified encoding scheme and fitness function, in order to put forward a more optimal scheme of warehouse layout, provide the information base for following operation of warehouse management, and improve the intelligent level of warehouse management system.

Abstract: Currently, rough set theory has been widely used in many fields. In rough set theory, how to measure the attributes significance of the data is a core content. In order to solve the problem that the existing attributes significance measure methods usually ignore the interaction among the attributes, the paper presents a measure method based on difference degree. When given a set, the proposed method first divides it into several subsets according to the value of condition attributes, and then computes the difference degree in the subsets. Secondly, the important attributes are selected based on the value of difference degree. Further the paper discussed some properties of the difference degree, and the experimental results shows the effectiveness of this method in the final.

Abstract: The paper proposes a new method that utilizes Parallel Estimation of Distribution (PED) to obtain disparity map of two images and fuses 3D point-cloud by the disparity. Estimation of Distribution (ED) has several advantages such as low complexity and high efficiency while it also has the shortcoming that it’s sensitive to initial samples and the final solution is local optimum but global optimum. In order to exert its merit and overcome the shortcoming, this paper will improve the ED by parallel sampling to diminish the sensitivity with Adaptive Support-Weight method. We called it Parallel Estimation of Distribution (PED). After disparity map is obtained, the two images will be divided into high frequency and low frequency by lifting wavelet respectively and the frequency coefficient of each image will be averaged into fusion frequency. Finally, the fused frequency coefficient will be transformed by inverse lifting wavelet to the final fused image. The experiment has proved that the matching speed is outstanding without losing precision.

Abstract: The genome compression algorithm based on the weighted context model is present in this paper. In this algorithm, the method to construct the weighted context model is discussed. The adaptive code length of the context models are used to determine the values of the corresponding weights of models. In the coding process, the values of these weights are modified in order to achieve the minimum code length and to enhance the compression efficiency. At last, the proposed algorithm is used to the bacteria genome compression. The experiments results indicate that the algorithm present could produce better compression result than the results by other algorithms.