Browsing by Author "Ingram, Branden Corwin"

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Emerging behaviours in artificial societies: simulating social-economic phenomena
    (2018) Ingram, Branden Corwin
    The behaviour of a society and its individuals is inherently complex and, therefore it becomes difficult when trying to model how changes in a society will affect that population. This dissertation presents an implementation of a computer simulation technique known as an Artificial Society, which is used to model social-economic phenomena using a multi-agent system. An Artificial Society is a system of simulated agents in a simulated world governed by a set of rules which handle the nature of the interactions between entities of the system. The purpose of this model is to analyse the emergence of global behaviours that form from the evolution of the society as a result of interactions governed by simple rules. We firstly expand on a number of aspects of the work done by Epstein and Axtell [1996] such as trade and cultural transmission. We analyse the similarities of our model with that of pre-existing models such as the Predator-Prey model. We demonstrate the spread of culture through a society and investigate the influence an individual can have on a population in different environments. We also investigate and analyse the benefits and shortfalls that different kinds of trade, taxation and investment can have on a society in a 3 resource environment. Through all of these experiments, we were able to demonstrate the emergence of complex behaviours which formed as a result of the interactions of individuals. This approach of modelling whereby we generate simple rules and observe the emergence of global behaviour is gives us the ability to tackle modelling complex behaviours where using a closed form solution would be impractical. It is hoped that this study will inform readers on the potentials and benefits of Artificial Societies.
  • No Thumbnail Available
    Item
    Play-style Identification and Player Modelling for Generating Tailored Advice in Video Games
    (University of the Witwatersrand, Johannesburg, 2023-09) Ingram, Branden Corwin; Rosman, Benjamin; Van Alten, Clint; Klein, Richard
    Recent advances in fields such as machine learning have enabled the development of systems that are able to achieve super-human performance on a number of domains, specifically in complex games such as Go and StarCraft. Based on these successes, it is reasonable to ask if these learned behaviours could be utilised to improve the performance of humans on the same tasks. However, the types of models used in these systems are typically not easily interpretable, and can not be directly used to improve the performance of a human. Additionally, humans tend to develop stylistic traits based on preference which aid in solving problems or competing at high levels. This thesis looks to address these difficulties by developing an end-to-end pipeline that can provide beneficial advice tailored to a player’s style in a video game setting. Towards this end, we demonstrate the ability to firstly cluster variable length multi-dimensional gameplay trajectories with respect to play-style in an unsupervised fashion. Secondly, we demonstrate the ability to learn to model an individual player’s actions during gameplay. Thirdly we demonstrate the ability to learn policies representative of all the play-styles identified with an environment. Finally, we demonstrate how the utilisation of these components can generate advice which is tailored to the individual’s style. This system would be particularly useful for improving tutorial systems that quickly become redundant lacking any personalisation. Additionally, this pipeline serves as a way for developers to garner insights on their player base which can be utilised for more informed decision-making on future feature releases and updates. For players, they gain a useful tool which can be utilised to learn how to play better as well identify as the characteristics of their gameplay as well as opponents. Furthermore, we contend that our approach has the potential to be employed in a broad range of learning domains.