How Does Learning Methods And The Membership Structure Affect The Organizational Decision Performance In Dynamic Environment?

With the fast development of economy and technology, the market environment we live in is changing rapidly. As the basic unit of work organization, the team enables organizations having high resilience, strong innovation and sustained vitalities. In the process of operation, the team often need to solve a series of repetitive decision-making problems. And the organization is facing some challenges just like deciding what kind of structure the team should be built and what kind of choice strategy to be used in order to solve these problems and achieve higher performance.Intuitively, when the environment changes, all kind of teams need to actively explore new knowledge to improve their adaptability. However, when the environment is changing constantly, the value of new knowledge generated by exploration will be eroded by the changing environment, moreover, team members have to pay the cost of exploration and innovation, so a greater degree of exploration not necessarily lead to more revenue. Therefore, we still need detailed discuss about which selection strategy can maximize the final benefits in the changing environment.We build three kind of team structure and develop a computational model based on the canonical multiarmed bandit model to respectively study how to balance exploration and exploitation in the process of choosing plans with the purpose of achieving the maximum earnings in the dynamic environment.After the comparison and analysis of the experimental results of the different teams, we have some main conclusions. First, the dynamic environment will erode the value of existing knowledge and new knowledge, so that the optimal balance of choice strategy for each team will be changed also. Second, when the environment changes occasionally, moderate exploration help the team to adapt to the environment quickly, but if the environment changes too often, the exploitation will be a better choice. Besides, if the team members can communicate with each other, they can exchange information quickly and find a better solution. But for the team without trust relationship, the team members will need a greater degree of exploration. Moreover, from the viewpoint of accumulated assets, the complex team built by heterogeneous members is the optimal team structure, but the process to establish such a such and the relations among members are too complicated, besides, the experiences exchanged by others may contain noisy information which affect their confidence and decision-making, so more research is needed to explore how to filter and remove noisy information.