Search

Considering the privacy leakage risk of data streams mining, a differential private Naive Bayes classification algorithm for data streams (DP-NB) is proposed in the paper. The Gaussian Bayesian model is firstly adopted and the methods of feature selection, cross-validation and weighting mechanism are used to improve the model and solve the problem of concept drift. Then noise perturbation is added to the model parameters and incremental update is presented to optimize the Gaussian classifier model. Theoretical analysis of the DP-NB algorithm proves that it achieves $\mathcal{E}-$ differential privacy. Finally, experimental results show that the DP-NB algorithm has better classification accuracy while satisfying privacy protection.

In order to suppress the circulating current and avoid the instantaneous short-circuit of the DC power, the high-value current-limiting inductors are needed between the cascaded units of the line-voltage cascaded voltage source inverter (LVC-VSI). In this paper, the three cascaded units of the LVC-VSI are considered as a whole, and the interaction between the switching states of the three units is analyzed. Besides, for the purpose of reducing the value of the current-limiting inductor, the constraints of switching combinations and the equivalent circuit under the condition of no circulating current and no short-circuit are deduced. Finally, taking voltage vector satisfying the constraints as the candidate vector, a finite control set model predictive control strategy is proposed. The experimental results show that the proposed strategy can realize good performance under the condition of low-value inductor.

Neutral-point (NP) voltage oscillation amplitude, THD of output voltage and the switching frequency of the inverter are the three key performance indices in three-level neutral-point-clamped (NPC) inverter. Every element connects with each other and restricts each other, which indicates that to keep the three key performance indices balanced is the research focus of the modulation strategy for three-level NPC inverter. Hybrid space vector PWM strategy, which is proposed in this paper, combines NV and non-NV, synthesizes the advantages of the two methods, reduces NP low-frequency oscillation and also decreases switching frequency and output voltage THD. Taking NTV (the most typical method of NV) and NTV (the most typical method of non-NV) as examples, A new hybrid SVPWM method, based on the NP current maximum minimum (MM) method is proposed in this paper. This new hybrid space vector PWM method applies NTV to uncontrolled balancing area and minimizes the switching frequency and THD of output voltage within the range of the allowable NP oscillation amplitude. Meanwhile, the impact on the output voltage produced by NP oscillation is researched in this paper. The experimental results show the validity of the proposed strategy.

To evaluate the pilots' safety behaviors for airline flight, an evaluation method Based on fuzzy neural network is investigated. Firstly, an evaluation index system of airline pilots' safety behaviors under the background of Chinese aviation safety culture is established. Secondly, for the first time, on the basis of combing fuzzy theory and neural network, the improved fuzzy neural network technology is applied to evaluate airline pilots' safety behaviors. Finally, a case study is put in practice, and simulation results through testing model show that the proposed method can evaluate airline pilots' safety behaviors effectively, also it will be very helpful for airline companies' choosing and training airline pilots in the future.

A highly efficient new fluoride absorbent was prepared by modification of sepiolite with an iron chloride. The effects of absorbent size, duration, pH, raw concentration, co-existing anions on removal efficiency of fluoride were investigated. The experiments indicated that Fe-Type Sepiolite is an effective fluoride absorbent with the removal efficiency above 95% by dosage of 2 g/100ml and duration of 40 min. The optimum range size of the absorbent is between the range of 0.–1.25 mm, and pH is partially sub acidity or neutral. Co-existing anions such as NO 3 − and SO 4 2− has a little Influence on adsorption of Fluoride. According to the analysis of adsorption isotherm and regeneration of Fe-Type Sepiolite, the mechanism of adsorbed fluoride is mainly by chemical adsorption.

Lines of waiting customers are always very long in most of banks. The essence of this phenomenon is the low efficiency of queuing system. In this paper, the queuing number, the service windows number and the optimal service rate are investigated by means of the queuing theory. In technology, the optimal problem of the bank queuing is solved. The time of customer queuing is reduced. The customer satisfaction is increased. It was proved that this optimal model of the queuing is feasible. By the example, the results are effective and practical.

The Internet represents a fundamentally different environment for businesses. It opens up the industry to increased competition and, as a result, reduces the likelihood of businesses retaining customers in the long run. So, it is important to analyze the relationship between customer values and customer retention in order to provide a guideline of successful e-business. The purpose of this paper is to answer the following questions. First, what are key components of customer value in e-business? Second, what influences the relationship between customer value and customer retention? Data collected from 520 effective respondents who made purchases through Internet shop were used to test a research model. Several managerial implications were derived from the analysis and further studies were suggested.

In practice, in purchasing decision-making, many quantitative and qualitative factors, with vagueness and imprecision, have to be considered. This makes the decision process very complicated and unstructured. Fuzzy sets theory is one of the methods used to deal with uncertain information. In this paper, a new approach based on fuzzy sets group decision is proposed to deal with the supplier selection problem in supply chain systems. The work procedure is shown briefly, as follows: First, linguistic values are used to assess the ratings and weights for quantitative or qualitative factors. Second, degree of similarity and probability of fuzzy sets are used to determine the ranking order of all alternatives. Finally, a numerical example of the selection problem of suppliers is shown, to highlight the procedure of the proposed approach, at the end of this paper.

Agro-food Marketing is a complex task involving various stack-holders, products and business scenarios. In a developing country like china, this activity is influenced by local, socio-economic and cultural characteristics. While other complex but well-defined business processes are experiencing benefits of services driven e-business; 'marketing of agro-food' has remained untouched by this revolution. Unavailability of proper underlying IT infrastructure will continue to inhibit the implementation and penetration of such technological advancements amongst the users. In absence of such capability, the informal transactions provide meager benefits to a farmer who looses a better price in other potential market, or a wholesaler, who might have got the desired quality product at a lower cost directly from the farm. There is a need to develop affordable and reliable solution that links all the actors involved in the system and provide an environment for a competitive business. This research paper demonstrates Agro-food E- Marketing Framework, developed to aid marketing of agro-food. With the help of the Agro-food E-Marketing Framework, a trader can execute a complete trading business process. This paper also demonstrates a complete business process that is initiated by a farmer with a mobile PDA involving trade of an agro-food in the market place. buyers and sellers perspectives, conventional markets are not flexible enough to provide business of choice. (2) provides effective case study explaining how a traveler can select services from a large pool of options to integrate appropriate business functions in planning a travel. It typically reveals the concept of integrating Business Processes to achieve a desired functionality. A business process typically involves user preference collection, product suggestion, price negotiation, trade agreement, authentication, fee calculation and collection, transaction management and other required tasks. Coming back to agriculture marketing scenario, if agriculture markets also follow standard business conventions all over the nation, then agro-food marketing can also be covered in integrated business services. Web Service Orchestration is an active area of research, involving the efforts to improve Web Services composition for business quality performance, quality of service and security in business services. Our objective is to test these technological advancements in a fairly complex, real life problem. This paper argues that if existing penetration of mobile communication amongst the farmers is harnessed properly, they also can be covered in the process of successful implementation of e-business. This paper proposes an effective Agro-food E-Marketing Framework to realize the expected outcomes.

An increasing number of people are processing transactions online and the numbers are likely to increase rapidly in the near future. It is important to analyze the relationship between customer values of Electronic Commerce (E-customer value) and Repurchase Intention in order to provide a guideline of successful e-business. The purpose of this paper is to answer the following questions. First, what are key components of customer value in e-business? Second, what influences the relationship between customer value and Repurchase Intention? Data collected from 350 respondents who made purchases through Internet shop were used to test a research model. Managerial implications were derived from the analysis and further studies were suggested.

Online trust is one of the key obstacles to vendors succeeding on the internet medium; a lack of trust is likely to discourage online consumers from participating in e-commerce. Previous study most have investigated only a component of e-commerce (e.g., initial adoption or continuance), but neglected the role of good relations with the consumer in ensuring a successful sustained relationship. Therefore, a model that went beyond intention and included key relational concepts (satisfaction, loyalty, etc.) was developed. Trust and its components are a major part of this model, which was based on strong theoretical foundations. Ten hypotheses were formulated. Data on the constructs were collected from 350 respondents and analyzed using lisrel 8.7 to test model validity and the hypotheses. Implications for researchers and practitioners are discussed.

An increasing number of people are processing transactions online and the numbers are likely to increase rapidly in the near future. This research focuses on the major considerations interacting with online bank and systematically measures factors affecting customer's interactions with online bank services. The results showed that trusting beliefs had the strongest relationship for users' intention to process transaction online. The second important affecting factor of interaction intention is perceived usefulness, following by ease of use and perceived Security. The findings in this study are useful for online bank vendors to understand what factors affect customer's interactions with online bank so as to acquire and retain customers.

One of the most important benefits of ANN (artificial neural networks) is the free of the restrictions of the classic statistical techniques. Some literatures points to several limitations in multiple regressions that are overcome by ANN. This paper demonstrates the usefulness of ANN in customer satisfaction analysis and compares ANN and regression, based on data from a China company customer satisfaction survey. Based on the results of this study, there is sufficient evidence to suggest that the application of ANN in customer satisfaction analysis is useful in identifying existing patterns in the data, and synergies between the drivers of satisfaction. The advantages of using ANN are highlighted and the managerial implications of ANN to identify the key drivers and set priorities for improvements are demonstrated.

Survival in a knowledge-based economy is derived from the ability to convert information to knowledge. To do so, researchers and managers are increasingly relying on the field of data mining. Data mining identifies and confirms relationships between explanatory and criterion variables. Data mining are increasingly popular because of the substantial contributions they can make in converting information to knowledge. Marketing is among the most frequent applications of the techniques, and whether you think about product development, advertising, distribution and retailing, or marketing research and business intelligence, data mining increasingly are being applied. The purpose of this paper is to provide an overview of data mining, summarize how it impact on knowledge creation in marketing, and suggest future developments in marketing and data mining for both organizations and researchers.

"Customer retention" is an increasingly pressing issue in today's ever-competitive commercial arena. This is especially relevant and important for sales and services related industries. This paper focus on constructing a new customer retention framework based on BP ANN (artificial neural network) and survival analysis. The new customer retention framework constituted of two parts, one is the BP ANN survival analysis model which integrates ANN technology and survival analysis to predict customer likelihood of defection and estimating customer value, the other is CLV (customer lifetime value) & CSP (customer survival phase) segment model which segment customer into different cluster to target high value, retainable customer. Effective actions triggered by these models could be the key to eventual customer retention. The empirical analysis is based on 23560 customers selected from the data warehouse of a large Chinese telecommunications company. This paper also suggests some propositions for further research.

"Customer Segmentation" is an increasingly pressing issue in today's over-competitive commercial arena. More and more literatures have researched the application of data mining technology in customer segmentation, and achieved sound effectives. But most of them segment customer only by single data mining technology from a special point, rather than from a systematical framework. Furthermore, one of the key purposes of customer segmentation is customer retention. Although previous segment methods may identify which group needs more care, it is unable to identify customer churn trend for adopting different action. This paper focus on proposing a customer segmentation framework based on data mining and constructs a new customer segmentation method based on customer survival character. The new customer segmentation method constituted of two steps. Firstly, using K-means clustering arithmetic cluster customer into different segments which has the similar survival function (churn trend) inside. Secondly, the method uses survival analysis to predict each cluster's survival/hazard function to test the validity of clustering and identify customer churn trend. The method mentioned above was applied in a dataset from China Telecommunications Company, which result proposes some useful management measures and suggestions. This paper also suggests some propositions for further research.