CadiaPlayer: A Simulation-Based General Game Player

Abstract. General Game Playing aims at AI systems that can understand the rules of new games and learn to play them effectively without human intervention. Our paper takes the first step towards general game-playing robots, which extend this capability to AI systems that play games in the real world. We develop a formal model for general games in physical environments and provide a systems architecture that allows the embedding of existing general game players as the "brain" and suitable robotic systems as the "body" of a general game-playing robot. We also report on an initial robot prototype that can understand the rules of arbitrary games and learns to play them in a fixed physical game environment.

Section: Prototypementioning

confidence: 99%

“…Established methods include Monte Carlo tree search [2], the automatic generation of evaluation functions [4] and knowledge acquisition [7]. Several complete general game-playing systems also have been described [12,2].…”

Section: Introductionmentioning

confidence: 99%

Towards General Game-Playing Robots: Models, Architecture and Game Controller

Rajaratnam

Thielscher

2013

“…Matsumoto et al [8] incorporated domain knowledge to guide the MC roll-outs, obtaining better results with a little more computational effort. Similarly, Björnsson and Finnsson [1] proposed a modification of standard UCT in order to include the best results of the simulations (in addition to average results) to drive the search towards the most promising regions. They also stored good lines of play found during the search which may be used effectively in single-player games.…”

Section: Literature Reviewmentioning

confidence: 99%

Monte-Carlo Tree Search for the Physical Travelling Salesman Problem

Pérez

Rohlfshagen

Lucas

2012

Abstract. The significant success of MCTS in recent years, particularly in the game Go, has led to the application of MCTS to numerous other domains. In an ongoing effort to better understand the performance of MCTS in open-ended realtime video games, we apply MCTS to the Physical Travelling Salesman Problem (PTSP). We discuss different approaches to tailor MCTS to this particular problem domain and subsequently identify and attempt to overcome some of the apparent shortcomings. Results show that suitable heuristics can boost the performance of MCTS significantly in this domain. However, visualisations of the search indicate that MCTS is currently seeking solutions in a rather greedy manner, and coercing it to balance short term and long term constraints for the PTSP remains an open problem.

“…In its basic form a wide range of games can be modeled, though there are severe restrictions, namely to deterministic games of full information, which immediately rule out most dice and card games. In recent years the most successful GGP agents such as CadiaPlayer [1] or Ary [11] made use of the UCT algorithm [8], which often delivers good results in multi-player games but lacks in single-player settings.…”

Section: Introductionmentioning

confidence: 99%

Translating Single-Player GDL into PDDL

Rauber

Kissmann

Hoffmann

2013

Abstract. In the single-agent case general game playing and action planning are two related topics, so that one might hope to use the established planners to improve the handling of general single-player games. However, both come with their own description language, GDL and PDDL, respectively. In this paper we propose a way to translate single-player games described in GDL to PDDL planning tasks and provide an evaluation on a wide range of single-player games, comparing the efficiency of grounding and solving the games in the translated and in the original format.