Research On The Game Model Based On AHP And Search Algorithms

Game theory as a cutting-edge areas of modern economics has become one of the main basic analysis tools. With the development and complement of game theory, game theory has been applied to many aspects of life. However, the technical analysis for the game is limited with a strict precondition, rigorous logic and clear thinking. In the analysis, the traditional solu-tions to the game problem have their own limitations, so it is necessary for us to combine with practical problems and to choose the suitable methods of solution. The use of AHP (analytic hierarchy process) to analyze static game including a variety of factors, and the use of hash technology andαâ†'βpruning search algorithm to solve complete information dynamic game, both are effective to solve the game problems.In this paper, the model of democracy appraisement in cadre management and the dots and boxes computer game system are to be the research topics. On the basis of reading and research the relevant literatures, we choose the way to analyze and to model according to the actual model. Again in accordance with the general principles of the artificial intelligence and computer game, the basic model of dots and boxes is analyzed from the perspective of game theory. The work includes the following aspects:Firstly, research on the game model of the cadre management and analysis for the model of democracy appraisement is done. Mainly, the bayesian game mode existing in the cadre man-agement is studied detailedly. Through discussion on the relationship between evaluation of the probability of management and the level of probability of the overall quality of cadres, the equi-librium of management's evaluation is given. AHP is used to build the model of AHP and the paired comparison matrix, and to do the consistency test. The combinative weight vectors are Calculated and the combinative consistency tests are done. By analyzing the eigenvector corre-sponding to the largest eigenvalue of the matrix, decision-makings are chose to solve practical problems. Finally, compared with actual results, the AHP model to solve the game problems is effective.Secondly, The theory of dots and boxes is researched. Based on the characteristics of this game, the key technologies of the game are analyzed, such as the long chain rule and the way of dealing with the incomplete states with a number of chains and cycles. The way of having greater probability for winning is well known. Knowledge of graph theory is used to explain the ways to deal with the mess in the game. And the results from the different boards are analyzed detailedly. Combining the characteristics of dots and boxes and using translation table for searching, the search of the game tree is made. The valuation of each state is given, according to the theorem.