Study On The Optimum Boundary Of Virtual Enterprises Based On K-Means Algorithm

Compared with the traditional business, virtual enterprises are more flexibility, adaptability and competitiveness, but due to many factors constraining, such as technology, economy, policies and so on, virtual enterprises still have to face the choice of members, as well as control the scale of the organization. Based on these realities, theory of virtual enterprises'boundary is proposed.From the perspective of the relationship theory and system theory, the influential factors of the virtual enterprises'boundary can be divided into two major categories: members and environment. The coordination and control ability, core resources and competencies, flexibility and compatibility of the organization are the main components of the members'aspect. Besides, the analysis of the environmental factors is only the market influence: size, cost and life cycle of the product of service .Based on the above analysis of this issue to the boundary of virtual enterprises, "Coupling" indicators is used to measure the interaction between members, the smaller the value of the coupling is, the higher the energy which produced by the cooperation between the members becomes; "entropy" measures the response capacity of the organization to the environment, the smeller margin means that the entropy of virtual enterprise organization is bigger. Integrated two indicators, when the sum value meets minimum, the virtual enterprises will get most energy.Take into accounting of the form process of the virtual enterprises, the issue choices K-Means algorithm and the Java language as tools, considering the economic significance and runtime costs, coupling is use to be the input of the K-Means algorithm, which includes consistency, shares, flexibility and complementarity. And the entropy which contains the three dimensions - size, cost and cycle is a control variable.