臺灣博碩士論文加值系統

猜拳是一種產生隨機結果以作決策的遊戲，有時生活中遇到一些無法決定的瑣事時，猜拳會是一個很常用的方法，但有時它並不隨機，因為玩家常常可以根據經驗判斷對方的手法，進而去預測對方出拳，猜拳本身也包含了個人的心理因素及反應靈敏度，只要能摸透對方猜拳的習慣，將可以有效提高自己的勝率。在本篇論文中，於iOS開發環境下撰寫猜拳遊戲App，針對22位不同的玩家總共進行了約6000 回合的對戰，並使用這些數據建構隱馬可夫模型，搭配維特比演算法決策最後結果。系統與玩家其中一方先獲勝120次即為贏家，一開始系統會先與玩家隨機猜拳20次，抓取玩家的出拳習慣作為基礎數據，訓練初始隱馬可夫模型，並在之後的每一輪都會把新資料加進去，達到即時更新數據庫，實驗結果證實了此方法能有效提高系統勝率。

Guessing is a game that produces random results to make decisions. Sometimes when encountering some undecidable trivial things in life, Guessing is a very common method, but sometimes it is not random, because players can often judge each other based on experience The method of predicting the opponent's punches, the guessing box itself also contains personal psychological factors and reaction sensitivity, as long as you can understand the opponent's boxing habits, you will be able to effectively improve your winning rate. In this paper, I wrote a boxing game App in the iOS development environment, and conducted a total of about 6000 rounds of battles against 22 different players. The data was used to construct a hidden Markov model, and the final result was decided by the Viterbi algorithm. One of the system and the player wins 120 times as the winner. At the beginning, the system will randomly guess the punch 20 times with the player, grab the player’s punching habits as the basic data, train the initial hidden Markov model, and use it in every subsequent round. New information is added to update the database in real time. The experimental results prove that this method can effectively improve the system win rate.

摘要 I
Abstract II
目錄 III
圖目錄 V
表目錄 VII
第一章 緒論 1
1.1 前言 1
1.2 研究動機 1
1.3 文獻回顧 1
1.4 論文架構 2
第二章 研究工具與演算法 3
2.1 iOS與Xcode 3
2.1.1 Swift 5
2.2 馬可夫鏈及其概念 6
2.2.1 馬可夫鏈模型 6
2.2.2 隱馬可夫模型 8
2.2.3 Viterbi演算法 11
2.3 使用者介面設計 14
第三章 系統實作 15
3.1 系統架構 15
3.2 系統建置 16
3.2.1 隱馬可夫模型結構 16
3.3 系統執行過程 18
第四章 實驗結果 20
4.1 使用者與電腦猜拳勝率 20
4.2 將模型緩衝器筆數改變 27
第五章 結論與未來展望 32
5.1 結論 32
5.2 未來展望 32
第六章 參考文獻 33

[1]東京石川奧研究室, http://ishikawa-vision.org/Booklet/index-j.html
[2]勝率高達100%的猜拳機器人,
https://designwork-s.net/gadgetproduct/superfast-rock-paper-scissors-robot.html
[3]Xcode, https://zh.wikipedia.org/wiki/Xcode
[4]朱克剛,” iOS 11程式設計實戰-Swift 4 快速上手的開發技巧200+”, 碁峰文化,2017
[5]猜拳遊戲介面設計, https://medium.com/%E5%BD%BC%E5%BE%97%E6%BD%98%E7%9A%84-swift-ios-app-%E9%96%8B%E7%99%BC%E6%95%99%E5%AE%A4/10-%E7%8C%9C%E6%8B%B3%E9%81%8A%E6%88%B2-%E5%89%AA%E5%88%80%E7%9F%B3%E9%A0%AD%E5%B8%83-b41203ba51c3
[6]Characteristic extraction method of human’s strategy in the Rock-Paper-Scissors game, Takahiro Komai1 , Song-Ju Kim2 , and Hiroaki Kurokawa1
[7]rock paper scissors game using Double-DQN against different random generator,https://github.com/dennylslee/rock-paper-scissors-DeepRL?fbclid=IwAR0y-nKSErfQz_GEXZIa7sfV29JuHoS7a0dXrFz7pByZln3RiY0fPZ3sJsY
[8]Mathias Zink, Paulina Friemann and Marco Ragni,,” Predictive Systems: The Game Rock-Paper-Scissors as an Example,” PRICAI 2019: Trends in Artificial Intelligence, PP. 514-526, 2019.
[9]Rock Paper Scissors Playing the Game with Markov Chains, Neil Pomerleau
[10]Swift5, https://developer.apple.com/cn/swift/
[11]Hidden Markov Models, https://web.stanford.edu/~jurafsky/slp3/A.pdf?fbclid=IwAR3p1Vkl6_NaP9FbY8sQXQMDWdS-w8_E--jv934ZALhhc1J_wmoOGW9fQw0
[12]隱馬可夫模型, https://blog.csdn.net/likelet/article/details/7056068
[13]Simon N, “iOS 11 App程式設計實戰心法：30個開發專業級iOS App的必學主題與實務講座”,博碩文化,2017
[14]Playing rock-paper-scissors between LSTM and random generators,https://github.com/dennylslee/rock-paper-scissors-LSTM?fbclid=IwAR1__VRAQN3uB8L1_UpE92u8g0zg63C2FJUs_NkEktXOQl63-jgmgxs6UO0
[15]How to win over 70% matches in Rock Paper Scissors,https://towardsdatascience.com/how-to-win-over-70-matches-in-rock-paper-scissors-3e17e67e0dab
[16]鄭捷,從Python入手+演算法,佳魁資訊股份有限公司,2017
[17]姜祖恕,馬可夫練的簡介,
http://episte.math.ntu.edu.tw/articles/mm/mm_09_3_08/index.html
[18]隱藏式馬可夫模型,
https://zh.wikipedia.org/wiki/%E9%9A%90%E9%A9%AC%E5%B0%94%E5%8F%AF%E5%A4%AB%E6%A8%A1%E5%9E%8B
[19]維特比演算法, https://kknews.cc/zh-tw/tech/5aek823.html
[20]iOS系統架構圖, https://medium.com/@wuufone/ios-%E6%A1%86%E6%9E%B6-%E6%9E%B6%E6%A7%8B%E5%9C%96%E6%95%B4%E7%90%86-411a73fff3ed