《有效讨论的最优混合解.doc》由会员分享，可在线阅读，更多相关《有效讨论的最优混合解.doc（2页珍藏版）》请在三一办公上搜索。
1、有效讨论的最优混合解 摘要公司为长远规划，会展开1些有成果性的讨论会议，但由于公司对小组讨论人员分配不均，导致有某些有权势的人影响优成果的讨论，所以平均分配小组讨论人员是很重要的。本论文要解决的问题就是把每个董事平均的分配到每段会话的讨论小组中。本文设定矩阵C（0，1）来记录设计好的组，用矩阵T来记录普遍知名时期的每两个成员的决策。在每个董事最佳参加会议次数为6的前提下，我们建造直角的拉丁矩阵来获得上午开会名单。至于下午的开会名单，我们采用Greedly（贪婪）算法思想。依照有计划的分配，我们设计出最佳的那个讨论组效率最高。这种小组分配方法能够广泛的应用于大中型集团公司，运用这种方法能够保证公
2、司、集团内部小组会议更加公正，避免了某些有权势的人支配会议。关键字:贪婪算法；分配名单；拉丁矩阵。AbstractCompany for farsighted program, will launch some have results to show a fruit discussion meeting, but because of company to group discussion the personnel dont all assign, causing to is some have the discussion that the shadow of human figure
3、of the power and influence rings an excellent result, so average allotment group discussion the personnel is very importantIn this paper,we solve the problem to design to assignments of several sessions with a different mix of people in each group .We use a( 0 ,1 )-matrix C to record designed groups
4、 ,and a matrix T to record currently known times of every two members discussions. After concluding that the optimal attending times of each member are 6, we construct orthogonal Latin Squares to obtain the morning list. For the afternoon list, we use the ideas of Greedy Algorithm. According to the
5、designed assignments, we design the current discussion group most efficiently.Furthermore, we perform computer simulation to indicate that the model makes sense. We summarize a set of general and practicable schemes. In generalization, we estimate the proper range by the empirical formular for a given number of members.The metod can be widely applied to experimental design.Key words :Greedy Algorithm；Assign list；Orthogonal Latin Square.
