• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.10a
• /
• pp.351-354
• /
• 2016

The competition of Go between AlphaGo and Lee Sedol attracted global interest leading AlphaGo to victory. The core function of AlphaGo is deep-learning system, studying by computer itself. Afterwards, the utilization of deep-learning system using artificial intelligence is said to be verified. Recently, the government passed the loT Act and developing its business model to promote loT. This study is on analyzing IoT business environment using deep-learning AI and constructing specialized business models.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.40 no.6
• /
• pp.76-86
• /
• 2003

We define a string stability heuristically and divide the board into group territory in computer Go. Elements of string stability are eye(E), eye-like(EL), special-eye(SE), extension-point(EX), liberty(L) and connection-point(CP). A string stability have 5 levels that are complete alive, alive, unsettled, danger and killed level. A group is made strings and link-points and have the territory. Territory division of a group is acquired by strings stability and link-points which are marym-mo, hankan, nalil-ja, and twokan between string and string. We compare our method with the result of evaluation of professional player. As a result, the mean error is 8.7.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.44 no.1
• /
• pp.49-55
• /
• 2007

This paper presents a method of score counting for computer Go that includes the consideration of stability, management of dead stones, and an algorithm for score counting. Thus, method for managing dead stones, filling all dames, and making additional moves is presented, along with a score-counting algorithm, where dames are defined as empty points that are not included in the area of a group, while additional moves are required for life when filling all the dames. In experiments using the final positions of 362 games, a mean error of 8.66, 5.96, and 4.15 was recorded for the score counting produced by the CGoban, HandTalk, and proposed methods, respectively. The proposed method was confirmed by experiments where it was success fully applied to the final positions.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.37 no.2
• /
• pp.21-30
• /
• 2000

If there exists a model that calculates the proper candidate position whenever the game of Go is in progress, it can be used for setting up the prototype of the candidate generation algorithm without using case-based reasoning. In this paper, we analyze Go through combinatorial game theory and on the basis of probability matrix (PM) showing the difference of the territory of the black and the white. We design and implement a candidate generation system(CGS) to find the candidates at a situation in Go. CGS designed in this paper can compute Influence power, safety, probability value(PV), and PM and then generate candidate positions for a present scene, once a stone is played at a scene. The basic strategy generates five candidates for the Present scene, and then chooses one with the highest PV. CGS generates the candidate which emphasizes more defence tactics than attack ones. In the opening game of computer Go, we can know that CGS which has no pattern is somewhat superior to NEMESIS which has the Joseki pattern.

• Journal of the Korea Convergence Society
• /
• v.7 no.4
• /
• pp.123-131
• /
• 2016

In this paper, we discussed learner's capability enhancement education suitable for the age of artificial intelligence (AI) using game analysis and archival research based on the 2016 Google Deepmind Challenge match between AI that possessed the finest deep neural networks and the master Baduk player that represented the best of the human minds. AlphaGo was a brilliant move that transcended the conventional wisdom of Baduk and introduced a new paradigm of Baduk. Lee Sedol defeated AlphaGo via the 'divine move and Great idea' that even AlphaGo could not have calculated. This was the triumph of human intuition and insights, which are deeply embedded in human nature as well as human courage and strength. Convergence science education that cultivates student abilities that can help them control machines in the age of AI must be in the direction of developing diverse human insights and positive spirits embedded in human nature not possessed by AI via implementing hearts-on experience and topophilia education obtained from the nature.

• The Journal of the Convergence on Culture Technology
• /
• v.4 no.2
• /
• pp.197-204
• /
• 2018

One of the hottest issues in our 21st century is AI. Just as the automation of manual labor has been achieved through the Industrial Revolution in the agricultural society, the intelligence information society has come through the SW Revolution in the information society. With the advent of Google 'Alpha Go', the computer has learned and predicted its own machine learning, and now the time has come for the computer to surpass the human, even to the world of Baduk, in other words, the computer. Machine learning ML (machine learning) is a field of artificial intelligence. Machine learning ML (machine learning) is a field of artificial intelligence, which means that AI technology is developed to allow the computer to learn by itself. The time has come when computers are beyond human beings. Many companies use machine learning, for example, to keep learning images on Facebook, and then telling them who they are. We also used a neural network to build an efficient energy usage model for Google's data center optimization. As another example, Microsoft's real-time interpretation model is a more sophisticated translation model as the language-related input data increases through translation learning. As machine learning has been increasingly used in many fields, we have to jump into the AI industry to move forward in our 21st century society.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.1
• /
• pp.163-173
• /
• 2011

In this paper, we propose a new paradigm of augmented reality board game environment and a portable game assistant(PGA) which can help gamers with strategy information. Previous AR board games consist of a private and public space. The public space provides rules of the game and shows the scene of game. And the gamers control game pieces in the public space. The previous games use the RFIDs for recognizing positions of the pieces, and the VR/AR environment for providing the scene of the game. However the RFIDs are expansive, and the VR/AR environment is inconvenient because it uses additional devices: the DataGlove, the digital pen, and the HMD. The proposed system recognizes positions of real pieces using the computer vision technique, and uses a monitor to provide dynamic effects. In the private space, previous systems provide entire screen of game and position of specific pieces, but cannot be controled the pieces by gamers. Therefore, in this system, we provide PGA that helps the user to plan of the strategy individually using universally mobile. The PGA helps to plan the strategy in the individual area, and to play easily in the side of the user's convenience.

• Journal for History of Mathematics
• /
• v.22 no.1
• /
• pp.53-74
• /
• 2009

Black-out Game(Lightout, Merlin Game, ${\sigma}$+Game) is an interesting game on the chessboard, when you click a button with black or white color, it changes color of itself and other buttons who shares edges. With this rule, we win the game when we have a chessboard with all same color after we click some of the buttons of it. Pretty much of research has been made on founding the winnable strategy for this type of game. In this paper, we first introduce a history of mathematical modeling on this game. Then we develop an algorithm to offer a winnable blackout game of any size. Our tools also show our new algorithm works. Finally, we show how we can use this game in mathematics education.

• Journal of Korea Multimedia Society
• /
• v.10 no.4
• /
• pp.483-491
• /
• 2007

This paper proposes a Situation Simulation Framework for implementing computer game and describes the possibility of analyzing game with unfolding situation. In last few years game systems have been evolving from the performance of physical engine, network traffic in the on line game to the representation of rendering physical phenomena. In computer game, a situation hierarchy structure which allows the designer for simulating high-level specifications of game structure. Logically simulated environment is created by defining situations and events based on hierarchy structure of the situation. We classify events into explicit event is occurred by user and implicit event is occurred by system. Our study defines the existence of objects is the most prevalent factor applied to any event in game world. The advantages of this approach are able to allow for providing the conceptual design for simulation game and analyzing the situation in the game world. And this method allows us to decrease the complexity of system design and abstraction modeling for the simulation game. Specially, the introduction of the definition of events allows us to approach game design in a structural manner rather than by their classification. The proposed method was implemented in the "Shooting BaDuk" among games.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.3
• /
• pp.141-148
• /
• 2022

Recently, reinforcement learning combined with deep neural network technology has achieved remarkable success in various fields such as board games such as Go and chess, computer games such as Atari and StartCraft, and robot object manipulation tasks. However, such deep reinforcement learning describes states, actions, and policies in vector representation. Therefore, the existing deep reinforcement learning has some limitations in generality and interpretability of the learned policy, and it is difficult to effectively incorporate domain knowledge into policy learning. On the other hand, dNL-RRL, a new relational reinforcement learning framework proposed to solve these problems, uses a kind of vector representation for sensor input data and lower-level motion control as in the existing deep reinforcement learning. However, for states, actions, and learned policies, It uses a relational representation with logic predicates and rules. In this paper, we present dNL-RRL-based policy learning for transportation mobile robots in a manufacturing environment. In particular, this study proposes a effective method to utilize the prior domain knowledge of human experts to improve the efficiency of relational reinforcement learning. Through various experiments, we demonstrate the performance improvement of the relational reinforcement learning by using domain knowledge as proposed in this paper.