Font Size: a A A

Jiuqi Intensive Learning Game Research And Multimedia Courseware Development

Posted on:2020-08-15Degree:MasterType:Thesis
Country:ChinaCandidate:S WangFull Text:PDF
GTID:2437330578451487Subject:The modern education technology
Abstract/Summary:PDF Full Text Request
JIU originated from Gannan Tibetan Autonomous Prefecture in Gansu Province and Aba Tibetan and Qiang Autonomous Prefecture in Sichuan Province.It is a unique folk chess art widely spread by Tibetans.Go,chess and other chess pieces that are well known to the public have made great progress in computer game research.AlphaGo and JueYi have made these types of Go more concerned.Compared with these kinds of chess,research on computer game of JIU is very rare.At present,the study of JIU games in the literature includes Bayesian network-based pattern extraction research,chess shape-based game strategy research,etc.Knowledge-based design state transfer strategy,limited to a small range of chess value is not intelligent.At the same time,the information education for JIU is also very scarce.As a traditional and outstanding national culture unique to Tibetans,the popularization of information education for JIU is an important means to protect and pass on JIU.In order to improve the game research level of JIU,this paper applies the deep reinforcement learning algorithm verified in Go to the game of JIU,and adds the time difference algorithm to improve the instant which to avoid the need to use super power to train.At the same time,the use of informational education means to design and develop a set of JIU multimedia teaching courseware,which provides a useful supplement for the game research and information education research of JIU.All of this paper is as follows:1.Collect existing chess same research and other game types of reinforcement learning materials,use C#to build UCT tree according to JIU characteristics.The time difference algorithm is used to optimize the JIU reinforcement learning update process.2.The built-in JIU reinforcement learning model is used to train,adjust the update weight of the time difference algorithms to optimize the JIU reinforcement learning model.The trained model is compared with the existing JIU game system and the human chess players.Tlearning he winning rate is tested and the learning effect of the JIU reinforcement model is evaluated.3.Developed a JIU information teaching multimedia courseware for the characteristics of teaching objects.Complete the teaching goal setting,the analysis of the teaching object characteristics,the teaching scene design and the teaching frame design and the teaching content design,and use some software to process the corresponding materials of the JIU multimedia teaching courseware.And use the courseware master developed JIU multimedia courseware scripts.Taking JIU game-type multimedia courseware as the starting point,improve the level of JIU education,and improve students' logical thinking ability and problem-solving ability in the form of games in the process of JIU teaching.
Keywords/Search Tags:Computer game, Reinforcement learning, UCT, Temporal-Difference Learning, Multimedia courseware
PDF Full Text Request
Related items