Creating programs that can play games such as chess, checkers, and backgammon, at a high level has long been a challenge and benchmark for AI. Computer game playing is arguably one of AI's biggest success stories. Several game playing systems developed in the past, such as Deep Blue, Chinook and TD-Gammon have demonstrated competitive play against the top human players. However, such systems are limited in that they play only one particular game and they typically must be supplied with game-specific knowledge. While their performance is impressive, it is difficult to determine if their success is due to generally applicable techniques or due to the human game analysis. A general game player is an agent capable of taking as input a description of a game's rules and proceeding to play without any subsequent human input. In doing so, the agent, rather than the human designer, is responsible for the domain analysis. Developing such a system requires the integration of several AI components, including theorem proving, feature discovery, heuristic search, and machine learning. In the general game playing scenario, the player agent is supplied with a game's rules in a formal language, prior to match play. This thesis contributes a collection of general methods for analyzing these game descriptions to improve performance. Prior work on automated domain analysis has focused on generating heuristic evaluation functions for use in search. The thesis builds upon this work by introducing a novel feature generation method. Also, I introduce a method for generating and comparing simple evaluation functions based on these features. I describe how more sophisticated evaluation functions can be generated through learning. Finally, this thesis demonstrates the utility of domain analysis in facilitating knowledge transfer between games for improved learning speed. The contributions are fully implemented with empirical results in the general game playing system. / text

No description available.

No description available.

No description available.

The goal of General Game Playing (GGP) is to develop a system, that is able to automatically play previously unseen games well, solely by being given the rules of the game. In contrast to traditional game playing programs, a general game player cannot be given game specific knowledge. Instead, the program has to discover this knowledge and use it for effectively playing the game well without human intervention. In this thesis, we present a such a program and general methods that solve a variety of knowledge discovery problems in GGP. Our main contributions are methods for the automatic construction of heuristic evaluation functions, the automated discovery of game structures, a system for proving properties of games, and symmetry detection and exploitation for general games.

The goal of General Game Playing (GGP) is to develop a system, that is able to automatically play previously unseen games well, solely by being given the rules of the game. In contrast to traditional game playing programs, a general game player cannot be given game specific knowledge. Instead, the program has to discover this knowledge and use it for effectively playing the game well without human intervention. In this thesis, we present a such a program and general methods that solve a variety of knowledge discovery problems in GGP. Our main contributions are methods for the automatic construction of heuristic evaluation functions, the automated discovery of game structures, a system for proving properties of games, and symmetry detection and exploitation for general games.:1. Introduction 2. Preliminaries 3. Components of Fluxplayer 4. Game Tree Search 5. Generating State Evaluation Functions 6. Distance Estimates for Fluents and States 7. Proving Properties of Games 8. Symmetry Detection 9. Related Work 10. Discussion

I detta arbete jämfördes ett LSTM-nätverk med ett feedforward-nätverk för generellt Atari 2600 spelande. Prestandan definierades som poängen agenten får för ett visst spel. Hypotesen var att LSTM skulle prestera minst lika bra som feedforward och förhoppningsvis mycket bättre. För att svara på frågeställningen skapades två olika agenter, en med ett LSTM-nätverk och en med ett feedforward-nätverk. Experimenten utfördes på Stella emulatorn med hjälp av ramverket the Arcade Learning Environment (ALE). Hänsyn togs till Machado råd om inställningar för användning av ALE och hur agenter borde tränas och evalueras samtidigt. Agenterna utvecklades med hjälp av en genetisk algoritm. Resultaten visade att LSTM var minst lika bra som feedforward men båda metoderna blev slagna av Machados metoder. Toppoängen i varje spel jämfördes med Granfelts arbete som har varit en utgångspunkt för detta arbete.

The use of the Internet has increased exponentially over the last decade. Individuals across all continents are progressively engaging in cyberspace interactions at work, in education, and for leisure. These online interactions, unconstrained by the limitations of corporeal reality, offer the potential for unique presentations of the self. The general aim of the research described in this thesis was to examine self-presentation in cyberspace. The research focused on MOOs, multi-user, text-based, user-extensible online environments, as a likely site for identity experimentation and play in cyberspace. Two studies are described. In the first quantitative study, 75 university students logged on to the front page of a social MOO where they selected a screen name, chose their gender, and provided a character description. As hypothesised, self-presentations were more likely to be based on actual identity rather than hoped for or feared selves. Contrary to expectation, little evidence was found of gender play. Self-presentations were typically positively biased and results suggested that players also perceived themselves more positively in the online context. Although sex and age were generally unrelated to self-presentation strategies, previous online experience, ethnicity, and personality profiles helped to explain self-presentation behaviour. / A qualitative study of a further 20 students in an educational MOO explored players understanding of their initial self-presentational choices and their management of these self-presentations over a 12-week period. Findings from the second study were consistent with the results from the first quantitative study and confirmed a strong desire for authentic self-presentation. Despite this emphasis on authenticity, the intention to play with identity was manifest in the form of selective self-disclosure, fantasy play, and exaggeration of traits. Participants also reported behaving in less inhibited ways online. A low incidence of gender play was noted. The overt identity play assumed by the cyberspace literature was not found in either study. Rather self-presentation in the online context appears to be governed by essentially similar processes to those that shape self-presentation in the offline world. The implications of the findings for teaching and learning, particularly for educators who want to use MOOs for identity experimentation, are discussed.

General game playing is an area of artificial intelligence which focuses on creating agents capable of playing many games from some class. The agents receive the rules just before the match and therefore cannot be specialized for each game. Deepstack is the first artificial intelligence to beat professional human players in heads-up no-limit Texas hold'em poker. While it is specialized for poker, at its core is a general algorithm for playing two-player zero-sum games with imperfect information - continual resolving. In this thesis we introduce a general version of continual resolving and compare its performance against Online Outcome Sampling Monte Carlo Counterfactual Regret Minimization in several games.

Computational creativity seeks, in part, to develop autonomous agents that exhibit creativity. Language is an ideal creative domain for studying computer agents due to its rich interconnectedness and immense space of possible combinations. This dissertation explores the design, testing, and theory of creative computer systems that write microfiction and play the board game Codenames. The designs of these systems are all similarly based on building up creative artifacts from the underlying structure of the relationships between words. A critical component of the creative process is the ability to evaluate the quality of creative output. Human and automated assessment of our creative systems' outputs yields insights into the challenge of automated creative evaluation. Those insights are formalized into a novel paradigm for designing creative systems and theoretical analyses of the properties of creative domains that facilitate evaluation.