DOCSLIB.ORG

Thesis for the Degree of Doctor

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Thesis for the Degree of Doctor Thesis for the Degree of Doctor DEEP LEARNING APPLIED TO GO-GAME: A SURVEY OF APPLICATIONS AND EXPERIMENTS 바둑에 적용된 깊은 학습: 응용 및 실험에 대한 조사 June 2017 Department of Digital Media Graduate School of Soongsil University Hoang Huu Duc Thesis for the Degree of Doctor DEEP LEARNING APPLIED TO GO-GAME: A SURVEY OF APPLICATIONS AND EXPERIMENTS 바둑에적용된 깊은 학습: 응용 및 실험에 대한 조사 June 2017 Department of Digital Media Graduate School of Soongsil University Hoang Huu Duc Thesis for the Degree of Doctor DEEP LEARNING APPLIED TO GO-GAME: A SURVEY OF APPLICATIONS AND EXPERIMENTS A thesis supervisor : Jung Keechul Thesis submitted in partial fulfillment of the requirements for the Degree of Doctor June 2017 Department of Digital Media Graduate School of Soongsil University Hoang Huu Duc To approve the submitted thesis for the Degree of Doctor by Hoang Huu Duc Thesis Committee Chair KYEOUNGSU OH (signature) Member LIM YOUNG HWAN (signature) Member KWANGJIN HONG (signature) Member KIRAK KIM (signature) Member KEECHUL JUNG (signature) June 2017 Graduate School of Soongsil University ACKNOWLEDGEMENT First of all, I’d like to say my deep grateful and special thanks to my advisor professor JUNG KEECHUL. My advisor Professor have commented and advised me to get the most optimal choices in my research orientation. I would like to express my gratitude for your encouraging my research and for supporting me to resolve facing difficulties through my studying and researching. Your advices always useful to me in finding the best ways in researching and they are priceless. Secondly, I would like to thank Prof. LIM YOUNG HWAN - the dean of our project. From the first time I’ve came to Korea until the end of my Ph.D course, you often support, advice, and encourage me to get best ways in most of my hardness situations in my study progress in Korea. I also profoundly thanks to my committee members: professor KYUNGSU OH- chairman, Prof. LIM YOUNG HWAN, Prof. KEECHUL JUNG, doctor HONG GWANG JIN, doctor RIRAK KIM for attending as my committee members. I would like to thank you for your dazzling comments and suggestions. I would especially like to HCI lab’s members. Many of you have supported me in study and research for my Ph.D. course. Be sides, I deeply thanks to my wife Nguyen Thi Quynh Trang, my big family for all of the sacrifices and supports that you’ve made in helping me while I have to live alone for Ph.D course far away. I would like to thank to all friends of mines who encourage me to complete my program. TABLE OF CONTENTS ABSTRACT IN ENGLISH ······························································· ix ABSTRACT IN KOREAN ······························································· xii CHAPTER 1: INTRODUCTION ..................................................... 1 1.1 BACKGROUND ............................................................................................................... 1 1.1.1 Machine Learning ................................................................................................. 2 1.1.2 Deep learning ........................................................................................................ 3 1.1.3 Convolutional Neural Networks ............................................................................ 5 1.1.4 Go-game ................................................................................................................ 5 1.2 RESEARCH MOTIVATION ................................................................................................ 7 1.3 PERSPECTIVE AND OVERVIEW ........................................................................................ 8 1.3.1 Machine Learning development ............................................................................ 8 1.3.2 Deep learning history ............................................................................................ 9 1.3.3 Contemporary machine learning ......................................................................... 17 1.4 CONTRIBUTIONS .......................................................................................................... 17 CHAPTER 2: GO-GAME AND ITS COMPLEXITY .................. 19 2.1 CHALLENGE ................................................................................................................ 19 2.2 APPLYING DEEP LEARNING INTO GO-GAME ................................................................. 20 2.3 GAME’S RULE SETS, POSSIBLE MOVES, AND LEGAL MOVES .......................................... 23 2.3.1 Go-game’s rule sets ............................................................................................. 23 2.3.1.1 Making sure that the game will come to an end-state: ................................................ 23 2.3.1.2 Deciding the winner of the game: ............................................................................... 23 2.3.1.3 Determining whether a group of stones is dead or alive at endgame states: ............... 24 2.3.1.4 Board sizes: ................................................................................................................ 24 2.3.1.5 The Stones: ................................................................................................................. 25 2.3.1.6 Playing a game: .......................................................................................................... 26 2.3.1.7 Territory: ..................................................................................................................... 28 2.3.1.8 Ko: .............................................................................................................................. 28 2.3.1.9 Superko:...................................................................................................................... 28 2.3.1.10 Eye: ........................................................................................................................... 29 2.3.1.11 Shoulder Hit: ............................................................................................................. 29 2.3.1.12 Chain: ....................................................................................................................... 29 2.3.1.13 Seki (mutual life): ..................................................................................................... 30 - i - 2.3.1.14 Suicide: ..................................................................................................................... 31 2.3.1.15 Komi: ........................................................................................................................ 31 2.3.1.16 Kosumi: .................................................................................................................... 32 2.3.1.17 Handicap: .................................................................................................................. 32 2.3.1.18 Goal: ......................................................................................................................... 32 2.3.1.19 Score: ........................................................................................................................ 33 2.3.1.20 Endgame state: .......................................................................................................... 34 2.3.2 Possible moves .................................................................................................... 34 2.3.3 Legal moves ........................................................................................................ 35 2.4 THE COMPLEXITY OF MOVES AND STRATEGIES ............................................................ 37 2.4.1 State-space complexity: ....................................................................................... 37 2.4.2 Game tree size: .................................................................................................... 37 2.4.3 Ply: ...................................................................................................................... 39 2.4.4 The complexity of Go-game moves .................................................................... 40 2.4.5 Comparison between Go-game and Chess. ......................................................... 41 2.4.6 The complexity of Go-game’s strategies. ............................................................ 45 2.4.7 Basic strategic: .................................................................................................... 45 2.4.8 Opening strategy: ................................................................................................ 47 2.4.9 Middle phase and endgame: ................................................................................ 48 2.5 HOT TREND IN GO-GAME RESEARCHING AND THE CHALLENGE. .................................. 49 2.6 THE REASON PEOPLE MUST CONCERN IN GO-GAME RESEARCH. ................................... 51 2.7 UNDERSTANDING ABOUT THE HUGE COMPLEXITY OF GO-GAME: ................................. 52 2.8 UNDERSTANDING HUGE NECESSARY RESOURCES ......................................................... 53 2.9 PROVIDE A FULLY OBSERVATION .................................................................................. 54 2.10 SHOWING EXPERIMENTS IN “DEEP LEARNING APPLIED TO GO-GAME” ....................... 54 CHAPTER 3: DEEP LEARNING APPLIED TO GO-GAME .... 55 3.1 WHY APPLYING DEEP LEARNING INTO GAME BUILDING. ............................................... 55 3.2 TYPICAL EVALUATION FUNCTION ................................................................................
Load more

Recommended publications
  • CSC321 Lecture 23: Go
    CSC321 Lecture 23: Go Roger Grosse Roger Grosse CSC321 Lecture 23: Go 1 / 22 Final Exam Monday, April 24, 7-10pm A-O: NR 25 P-Z: ZZ VLAD Covers all lectures, tutorials, homeworks, and programming assignments 1/3 from the first half, 2/3 from the second half If there's a question on this lecture, it will be easy Emphasis on concepts covered in multiple of the above Similar in format and difficulty to the midterm, but about 3x longer Practice exams will be posted Roger Grosse CSC321 Lecture 23: Go 2 / 22 Overview Most of the problem domains we've discussed so far were natural application areas for deep learning (e.g. vision, language) We know they can be done on a neural architecture (i.e. the human brain) The predictions are inherently ambiguous, so we need to find statistical structure Board games are a classic AI domain which relied heavily on sophisticated search techniques with a little bit of machine learning Full observations, deterministic environment | why would we need uncertainty? This lecture is about AlphaGo, DeepMind's Go playing system which took the world by storm in 2016 by defeating the human Go champion Lee Sedol Roger Grosse CSC321 Lecture 23: Go 3 / 22 Overview Some milestones in computer game playing: 1949 | Claude Shannon proposes the idea of game tree search, explaining how games could be solved algorithmically in principle 1951 | Alan Turing writes a chess program that he executes by hand 1956 | Arthur Samuel writes a program that plays checkers better than he does 1968 | An algorithm defeats human novices at Go 1992
    [Show full text]
  • E-Book? Začala Hra
    Nejlepší strategie všech dob 1 www.yohama.cz Nejlepší strategie všech dob Obsah OBSAH DESKOVÁ HRA GO ................................................................................................................................ 3 Úvod ............................................................................................................................................................ 3 Co je to go ................................................................................................................................................... 4 Proč hrát go ................................................................................................................................................. 7 Kde získat go ................................................................................................................................................ 9 Jak začít hrát go ......................................................................................................................................... 10 Kde a s kým hrát go ................................................................................................................................... 12 Jak se zlepšit v go ...................................................................................................................................... 14 Handicapová hra ....................................................................................................................................... 15 Výkonnostní třídy v go a rating hráčů .......................................................................................................
    [Show full text]
  • FUEGO—An Open-Source Framework for Board Games and Go Engine Based on Monte Carlo Tree Search Markus Enzenberger, Martin Müller, Broderick Arneson, and Richard Segal
    IEEE TRANSACTIONS ON COMPUTATIONAL INTELLIGENCE AND AI IN GAMES, VOL. 2, NO. 4, DECEMBER 2010 259 FUEGO—An Open-Source Framework for Board Games and Go Engine Based on Monte Carlo Tree Search Markus Enzenberger, Martin Müller, Broderick Arneson, and Richard Segal Abstract—FUEGO is both an open-source software framework with new algorithms that would not be possible otherwise be- and a state-of-the-art program that plays the game of Go. The cause of the cost of implementing a complete state-of-the-art framework supports developing game engines for full-information system. Successful previous examples show the importance of two-player board games, and is used successfully in a substantial open-source software: GNU GO [19] provided the first open- number of projects. The FUEGO Go program became the first program to win a game against a top professional player in 9 9 source Go program with a strength approaching that of the best Go. It has won a number of strong tournaments against other classical programs. It has had a huge impact, attracting dozens programs, and is competitive for 19 19 as well. This paper gives of researchers and hundreds of hobbyists. In Chess, programs an overview of the development and current state of the UEGO F such as GNU CHESS,CRAFTY, and FRUIT have popularized in- project. It describes the reusable components of the software framework and specific algorithms used in the Go engine. novative ideas and provided reference implementations. To give one more example, in the field of domain-independent planning, Index Terms—Computer game playing, Computer Go, FUEGO, systems with publicly available source code such as Hoffmann’s man-machine matches, Monte Carlo tree search, open source soft- ware, software frameworks.
    [Show full text]
  • Move Prediction in Gomoku Using Deep Learning
    VW<RXWK$FDGHPLF$QQXDO&RQIHUHQFHRI&KLQHVH$VVRFLDWLRQRI$XWRPDWLRQ :XKDQ&KLQD1RYHPEHU Move Prediction in Gomoku Using Deep Learning Kun Shao, Dongbin Zhao, Zhentao Tang, and Yuanheng Zhu The State Key Laboratory of Management and Control for Complex Systems, Institute of Automation, Chinese Academy of Sciences Beijing 100190, China Emails: [email protected]; [email protected]; [email protected]; [email protected] Abstract—The Gomoku board game is a longstanding chal- lenge for artificial intelligence research. With the development of deep learning, move prediction can help to promote the intelli- gence of board game agents as proven in AlphaGo. Following this idea, we train deep convolutional neural networks by supervised learning to predict the moves made by expert Gomoku players from RenjuNet dataset. We put forward a number of deep neural networks with different architectures and different hyper- parameters to solve this problem. With only the board state as the input, the proposed deep convolutional neural networks are able to recognize some special features of Gomoku and select the most likely next move. The final neural network achieves the accuracy of move prediction of about 42% on the RenjuNet dataset, which reaches the level of expert Gomoku players. In addition, it is promising to generate strong Gomoku agents of human-level with the move prediction as a guide. Keywords—Gomoku; move prediction; deep learning; deep convo- lutional network Fig. 1. Example of positions from a game of Gomoku after 58 moves have passed. It can be seen that the white win the game at last. I.
    [Show full text]
  • Strategy Game Programming Projects
    STRATEGY GAME PROGRAMMING PROJECTS Timothy Huang Department of Mathematics and Computer Science Middlebury College Middlebury, VT 05753 (802) 443-2431 [email protected] ABSTRACT In this paper, we show how programming projects centered around the design and construction of computer players for strategy games can play a meaningful role in the educational process, both in and out of the classroom. We describe several game-related projects undertaken by the author in a variety of pedagogical situations, including introductory and advanced courses as well as independent and collaborative research projects. These projects help students to understand and develop advanced data structures and algorithms, to manage and contribute to a large body of existing code, to learn how to work within a team, and to explore areas at the frontier of computer science research. 1. INTRODUCTION In this paper, we show how programming projects centered around the design and construction of computer players for strategy games can play a meaningful role in the educational process, both in and out of the classroom. We describe several game-related projects undertaken by the author in a variety of pedagogical situations, including introductory and advanced courses as well as independent and collaborative research projects. These projects help students to understand and develop advanced data structures and algorithms, to manage and contribute to a large body of existing code, to learn how to work within a team, and to explore areas at the frontier of computer science research. We primarily consider traditional board games such as chess, checkers, backgammon, go, Connect-4, Othello, and Scrabble. Other types of games, including video games and real-time strategy games, might also lead to worthwhile programming projects.
    [Show full text]
  • Collaborative Go
    Distributed Computing Master Thesis COLLABORATIVE GO Alex Hugger December 13, 2011 Supervisor: S. Welten Prof. Dr. R. Wattenhofer Distributed Computing Group Computer Engineering and Networks Laboratory ETH Zürich Abstract The game of Go is getting more and more popular in Europe. Its simple rules and the huge complexity for computer programs make it very interesting for a study about collaborative gaming. The idea of collaborative Go is to play a simple two player game in a setup of teams behaving as one single player instead of one versus one. During this thesis, we implemented a framework allowing to play collaborative Go. The framework is heavily based on the existing Go Text Protocol allowing an inte- gration into already existing Go servers or a competition with existing Go programs. Multiple players are merged through the use of different decision engines into one single player. Two decision engines were implemented based on different interac- tion schemes. One is based on a voting process whereas the other one decides on the final move of a team by applying a Monte Carlo simulation. The final experiments show that collaborative gaming is very interesting and atten- tion attracting for the players. Especially the voting mechanism used in the collab- orative player achieved a high user satisfaction. For an automated decision engine, we found out that it is very important to make the decision process as transparent as possible to all users, as this increases the level of the acceptance even when a bad decision was made. Acknowledgments During the accomplishment of this master thesis, a lot of people supported and guided me to a successful completion of the work.
    [Show full text]
  • Combining Tactical Search and Deep Learning in the Game of Go
    Combining tactical search and deep learning in the game of Go Tristan Cazenave PSL-Universite´ Paris-Dauphine, LAMSADE CNRS UMR 7243, Paris, France [email protected] Abstract Elaborated search algorithms have been developed to solve tactical problems in the game of Go such as capture problems In this paper we experiment with a Deep Convo- [Cazenave, 2003] or life and death problems [Kishimoto and lutional Neural Network for the game of Go. We Muller,¨ 2005]. In this paper we propose to combine tactical show that even if it leads to strong play, it has search algorithms with deep learning. weaknesses at tactical search. We propose to com- Other recent works combine symbolic and deep learn- bine tactical search with Deep Learning to improve ing approaches. For example in image surveillance systems Golois, the resulting Go program. A related work [Maynord et al., 2016] or in systems that combine reasoning is AlphaGo, it combines tactical search with Deep with visual processing [Aditya et al., 2015]. Learning giving as input to the network the results The next section presents our deep learning architecture. of ladders. We propose to extend this further to The third section presents tactical search in the game of Go. other kind of tactical search such as life and death The fourth section details experimental results. search. 2 Deep Learning 1 Introduction In the design of our network we follow previous work [Mad- Deep Learning has been recently used with a lot of success dison et al., 2014; Tian and Zhu, 2015]. Our network is fully in multiple different artificial intelligence tasks.
    [Show full text]
  • (CMPUT) 455 Search, Knowledge, and Simulations
    Computing Science (CMPUT) 455 Search, Knowledge, and Simulations James Wright Department of Computing Science University of Alberta [email protected] Winter 2021 1 455 Today - Lecture 22 • AlphaGo - overview and early versions • Coursework • Work on Assignment 4 • Reading: AlphaGo Zero paper 2 AlphaGo Introduction • High-level overview • History of DeepMind and AlphaGo • AlphaGo components and versions • Performance measurements • Games against humans • Impact, limitations, other applications, future 3 About DeepMind • Founded 2010 as a startup company • Bought by Google in 2014 • Based in London, UK, Edmonton (from 2017), Montreal, Paris • Expertise in Reinforcement Learning, deep learning and search 4 DeepMind and AlphaGo • A DeepMind team developed AlphaGo 2014-17 • Result: Massive advance in playing strength of Go programs • Before AlphaGo: programs about 3 levels below best humans • AlphaGo/Alpha Zero: far surpassed human skill in Go • Now: AlphaGo is retired • Now: Many other super-strong programs, including open source Image source: • https://www.nature.com All are based on AlphaGo, Alpha Zero ideas 5 DeepMind and UAlberta • UAlberta has deep connections • Faculty who work part-time or on leave at DeepMind • Rich Sutton, Michael Bowling, Patrick Pilarski, Csaba Szepesvari (all part time) • Many of our former students and postdocs work at DeepMind • David Silver - UofA PhD, designer of AlphaGo, lead of the DeepMind RL and AlphaGo teams • Aja Huang - UofA postdoc, main AlphaGo programmer • Many from the computer Poker group
    [Show full text]
  • Deep Learning for Go
    Research Collection Bachelor Thesis Deep Learning for Go Author(s): Rosenthal, Jonathan Publication Date: 2016 Permanent Link: https://doi.org/10.3929/ethz-a-010891076 Rights / License: In Copyright - Non-Commercial Use Permitted This page was generated automatically upon download from the ETH Zurich Research Collection. For more information please consult the Terms of use. ETH Library Deep Learning for Go Bachelor Thesis Jonathan Rosenthal June 18, 2016 Advisors: Prof. Dr. Thomas Hofmann, Dr. Martin Jaggi, Yannic Kilcher Department of Computer Science, ETH Z¨urich Contents Contentsi 1 Introduction1 1.1 Overview of Thesis......................... 1 1.2 Jorogo Code Base and Contributions............... 2 2 The Game of Go5 2.1 Etymology and History ...................... 5 2.2 Rules of Go ............................. 6 2.2.1 Game Terms......................... 6 2.2.2 Legal Moves......................... 7 2.2.3 Scoring............................ 8 3 Computer Chess vs Computer Go 11 3.1 Available Resources......................... 11 3.2 Open Source Projects........................ 11 3.3 Branching Factor .......................... 12 3.4 Evaluation Heuristics........................ 13 3.5 Machine Learning.......................... 13 4 Search Algorithms 15 4.1 Minimax............................... 15 4.1.1 Iterative Deepening .................... 16 4.2 Alpha-Beta.............................. 17 4.2.1 Move Ordering Analysis.................. 18 4.2.2 History Heuristic...................... 18 4.3 Probability Based Search...................... 19 4.3.1 Performance......................... 20 4.4 Monte-Carlo Tree Search...................... 21 5 Evaluation Algorithms 23 i Contents 5.1 Heuristic Based........................... 23 5.1.1 Bouzy’s Algorithm..................... 24 5.2 Final Positions............................ 24 5.3 Monte-Carlo............................. 25 5.4 Neural Network........................... 25 6 Neural Network Architectures 27 6.1 Networks in Giraffe........................
    [Show full text]
  • SAI: a Sensible Artificial Intelligence That Targets High Scores in Go
    SAI: a Sensible Artificial Intelligence that Targets High Scores in Go F. Morandin,1 G. Amato,2 M. Fantozzi, R. Gini,3 C. Metta,4 M. Parton,2 1Universita` di Parma, Italy, 2Universita` di Chieti–Pescara, Italy, 3Agenzia Regionale di Sanita` della Toscana, Italy, 4Universita` di Firenze, Italy [email protected], [email protected], [email protected], [email protected], [email protected], [email protected] Abstract players. Moreover, one single self-play training game is used to train 41 winrates, which is not a robust approach. Finally, We integrate into the MCTS – policy iteration learning pipeline in (Wu 2019, KataGo) the author introduces a heavily modi- of AlphaGo Zero a framework aimed at targeting high scores fied implementation of AlphaGo Zero, which includes score in any game with a score. Training on 9×9 Go produces a superhuman Go player. We develop a family of agents that estimation, among many innovative features. The value to be can target high scores, recover from very severe disadvan- maximized is then a linear combination of winrate and expec- tage against weak opponents, and avoid suboptimal moves. tation of a nonlinear function of the score. This approach has Traning on 19×19 Go is underway with promising results. A yielded a extraordinarily strong player in 19×19 Go, which multi-game SAI has been implemented and an Othello run is is available as an open source software. Notably, KataGo is ongoing. designed specifically for the game of Go. In this paper we summarise the framework called Sensi- ble Artificial Intelligence (SAI), which has been introduced 1 Introduction to address the above-mentioned issues in any game where The game of Go has been a landmark challenge for AI re- victory is determined by score.
    [Show full text]
  • Move Prediction in the Game of Go
    Move Prediction in the Game of Go A Thesis presented by Brett Alexander Harrison to Computer Science in partial fulfillment of the honors requirements for the degree of Bachelor of Arts Harvard College Cambridge, Massachusetts April 1, 2010 Abstract As a direct result of artificial intelligence research, computers are now expert players in a variety of popular games, including Checkers, Chess, Othello (Reversi), and Backgammon. Yet one game continues to elude the efforts of computer scientists: Go, also known as Igo in Japan, Weiqi in China, and Baduk in Korea. Due in part to the strategic complexity of Go and the sheer number of moves available to each player, most typical game-playing algorithms have failed to be effective with Go. Even state-of-the-art computer Go programs are weaker than high-ranking amateurs. Thus Go provides the perfect framework for devel- oping and testing new ideas in a variety of fields in computer science, including algorithms, computational game theory, and machine learning. In this thesis, we explore the problem of move prediction in the game of Go. The move prediction problem asks how we can build a program which trains on Go games in order to predict the moves of professional and high-ranking amateur players in other Go games. An accurate move predictor could serve as a powerful component of Go-playing programs, since it can be used to reduce the branching factor in game tree search and can be used as an effective move ordering heuristic. Our first main contribution to this field is the creation of a novel move prediction system, based on a naive Bayes model, which builds upon the work of several previous move prediction systems.
    [Show full text]
  • What Is the Project Title
    Artificial Intelligence for Go CIS 499 Senior Design Project Final Report Kristen Ying Advisors: Dr. Badler and Dr. Likhachev University of Pennsylvania PROJECT ABSTRACT Go is an ancient board game, originating in China – where it is known as weiqi - before 600 B.C. It spread to Japan (where it is also known as Go), to Korea as baduk, and much more recently to Europe and America [American Go Association]. This game, in which players take turn placing stones on a 19x19 board in an attempt to capture territory, is polynomial-space hard [Reyzin]. Checkers is also polynomial-space hard, and is a solved problem [Schaeffer]. However, the sheer magnitude of options (and for intelligent algorithms, sheer number of possible strategies) makes it infeasible to create even passable novice AI with an exhaustive pure brute-force approach on modern hardware. The board is 19 x19, which, barring illegal moves, allows for about 10171 possible ways to execute a game. This number is about 1081 times greater than the believed number of elementary particles contained in the known universe [Keim]. Previously investigated approaches to creating AI for Go include human-like models such as pattern recognizers and machine learning [Burmeister and Wiles]. There are a number of challenges to these approaches, however. Machine learning algorithms can be very sensitive to the data and style of play with which they are trained, and approaches modeling human thought can be very sensitive to the designer‟s perception of how one thinks during a game. Professional Go players seem to have a very intuitive sense of how to play, which is difficult to model.
    [Show full text]