Abstract:A vehicle routing and dispatching problem in a dynamic traffic network with real-time traffic information is discussed in this paper．Both re-current and non-recurrent congestion are considered in the problem．The initial route planning is made by solving a time-dependent vehicle routing problem under re-current congestion conditions．The genetic algorithm is applied．During the planning implementation，a new mechanism updating at the critical node is proposed to adjust the routes with real-time traffic information subject to the un_x005fexpected traffic accidents in non-recurrent congestion environment．It is conducted by solving a special traveling salesman problem with Dynasearch algorithm．The numerical simulation results with the Sioux Falls network show the efficiency of the new mechanism．The new strategy can improve the other strategies in a short time，and it becomes more efficient when the network becomes more dynamic．

Abstract:The extant literature suggests that entrepreneurial speed is the antecedent to new venture creation．However，there is little knowledge about the antecedents to entrepreneurial speed．This paper addressed the relationship between entrepreneurs’affiliation of prior experience and entrepreneurial speed，as well as the moderation effects of marketization and the affiliation of mobilized network contacts．The results showed that both prior experiences embedded in planned and market system could promote the entrepreneurial speed．Furthermore，in the context of higher marketization，the positive relationship between prior experience affiliated by planned system and entrepreneurial speed is higher than the context of lower marketization．And entrepreneur from market system could promote the entrepreneurial speed by mobilizing the network contact embedded in planned system．The results provide significant implications to the research on entrepreneurial behavior，entrepreneurial management and policy making．

Abstract:The prevalence of Internet，online consumer reviews and seller’s frequent discount strategies cantrain consumers’strategic waiting behavior，which deteriorates product mismatch between supply and demand． On the other hand，the“insurance mechanism”behind the return strategy via improving consumers’willingness to pay for product encourages more consumers to buy products at the regular price，which in turn results in plenty of returned products．How to deal with these returned goods has become a core problem the seller must face．This paper analyze how the three reprocess strategies of returned goods，i.e.，non-resale，regular resale，and discount resale strategy，influence the seller’s return strategy design．Our conclusion shows that: compared with the no return strategy，1) The non resale and discount resale strategies decrease the seller’s expected profits; 2) When both regular reprocess cost and refund are small，the regular resale strategy could increase the seller’s expected profits，however，if either regular reprocess cost or refund is large，the regular resale strategy is also detrimental to the seller’s profitability．

Abstract:The contract-based supply chain ( SC) coordination problem of a two-stage SC under random yield and random demand is studied．In the addressed SC，the decision variables of the supplier and the producer are the agricultural input ( I) and the order quantity of raw material ( R) ，respectively．The uniqueness of the SC is that both the producer and the supplier are characterized by a random yield． For the random proportional yield，firstly the concavity of the expected profit function of the whole SC with respect to I and R is proved．Then，we show that the revenue-sharing ( RS) contract is not able to coordinate the chain． Lastly，a RS and yield risk-sharing contract is presented to coordinate the chain．The proposed contract is based on RS contract and bidirectional compensation for overproduction and shortage of raw material to share the risk of supply uncertainty．The coordination effectiveness of the proposed contract is theoretically proved，and is illustrated by two numerical examples． The results of numerical analysis show that the expected profit of the producer is more sensitive to the change of supply and demand uncertainty than that of the supplier．

Abstract:Based on the fact that，in the actual project management，actions of the contractor and the client are interacted with each other and their profits are related closely，this paper investigates the project scheduling problem considering the interaction between the two parties of a contract． The authors identify the studied problem firstly，where the client exerts an effect on the contractor through the payment arrangement so as to obtain a desirable profit，while the contractor reacts to the client by adjusting activity schedule with the objective of maximizing his own profit．Then，taking the maximization of the project’s net present value as the objective，the authors construct a project scheduling optimization model which is composed of two submodels， namely，the contractor and the client submodel．For the strong NP-hardness of the problem，a special tabu search heuristic algorithm is developed． The algorithm is tested on a standard instance set generated randomly and the results show that the algorithm is an efficient algorithm for the problem．Ultimately，an example is utilized to illustrate the research and the following conclusions are drawn: With the postponement of the project deadline or the augment of the contractor’s profit baseline，the profit of the client tends to decrease while that of the contractor tends to increase．

Abstract:Currently，many studies dedicated to context aware based recommendation，considered different types of context properties，but they ignore the important degree of different context attribute impact the behavior，that is，context strength．This paper defines the customer context，context intensity，and behavior changing quantitatively; presents context strength constrained pattern mining methods and change detecting method to extract the critical situation caused by changes in behavior．The proposed algorithm increased the sensitivity of the interests changing，improvement of the massive data under support for sparse association rules，the shortcomings of low confidence sensitivity．Experiments and analysis demonstrated the feasibility and effectiveness．

Abstract:Considering the negative correlation of stock returns and its volatility，this paper established a timevarying infinite pure jump Levy processes with time-changed conditional expectations and volatility in discretetime．According to local martingale measure transformation，we derived its equivalent risk neutral pricing model for the conditional Levy processes and used in the Hang Seng Index options for empirical research．Studies show: the conditional Levy processes with leverage effect jointly portray the asset prices’time-varying drift， variance，non-Gauss random innovations and asymmetric volatility four states，and this model has wide applicability; compared to Brownian Motion，Jump-Diffusion，and Variance Gamma process，Tempered Stable models have better performance in capturing leptokurtosis and fat-tailed features; with leverage effect，option pricing capacity of conditional Levy process has been greatly improved，we also found that Rapidly Decreasing Tempered Stable process performs more robust．