General Game Playing with Imperfect Information

Abstract

General Game Playing is a field which allows the researcher to investigate techniques that might eventually be used in an agent capable of Artificial General Intelligence. Game playing presents a controlled environment in which to evaluate AI techniques, and so we have seen an increase in interest in this field of research. Games of imperfect information offer the researcher an additional challenge in terms of complexity over games with perfect information. In this article, we look at imperfect-information games: their expression, their complexity, and the additional demands of their players. We consider the problems of working with imperfect information and introduce a technique called HyperPlay, for efficiently sampling very large information sets, and present a formalism together with pseudo code so that others may implement it. We examine the design choices for the technique, show its soundness and completeness then provide some experimental results and demonstrate the use of the technique in a variety of imperfect-information games, revealing its strengths, weaknesses, and its efficiency against randomly generating samples. Improving the technique, we present HyperPlay-II, capable of correctly valuing information-gathering moves. Again, we provide some experimental results and demonstrate the use of the new technique revealing its strengths, weaknesses and its limitations.

Language for Description of Worlds. Part 1: Theoretical Foundation

“…Note: Various authors may use different terminology. Thus, the term used in [9] is "game" instead of "world". Likewise, [9] uses "Game Description Language" instead of "Language for Description of Worlds" as well as "Imperfect Information" instead of "Partial Observability".…”

“…Thus, the term used in [9] is "game" instead of "world". Likewise, [9] uses "Game Description Language" instead of "Language for Description of Worlds" as well as "Imperfect Information" instead of "Partial Observability". We can imagine the world as a game and, vice versa, a game can be thought of as a world in its own right.…”

Development of Answer Validation System Using Responders’ Attributes and Crowd Ranking

“…Different questions requiring correct answers are uploaded on the internet on daily basis which leads to the development of question answering (QA) systems, with the aim of providing accurate answers to explicit questions which are contrasting to document retrieval (Ojokoh & Journal of Service Science and Management Adebisi, 2019; Toba et al, 2014). Schofield and Thielscher (2019) defined community QA as a website or service that requires a method to display pieces of information in the form of a question in natural language, a medium for communal response and a community in which questions and answers are rooted based on the level of participation, and answers provided was discovered to be of higher quality when it was compared with other types of online QA services (Harper et al, 2008). However, answers to questions from users form the pillar of a successful CQA service, in which better answers may be provided as against automatic systems.…”

“…Several studies have been carried out on how to make better the quality of the answers provided by QA system, focusing on textual entailment, question type analysis, answer ranking by the crowd workers and domain experts and personal and community features (past history) of the answerer to determine the quality of the answers (Ríos-Gaona et al, 2012;Su et al, 2007;Ishikawa et al, 2011;Ojokoh & Ayokunle, 2012;Anderson et al, 2012;Schofield & Thielscher, 2019).…”

Learning state-action features for general game playing

“…When games are provided in a language such as L-GDL, it is not feasible to automatically infer any such knowledge or representations as required by these algorithms. The HyperPlay-II search algorithm (Schofield and Thielscher, 2019) is a rare exception that has no need for such domain knowledge, but it also has no clear integration with learning algorithms, and its backtracking mechanism for finding models that are consistent with the current history can be inefficient in long games with deep game trees.…”