1 Introduction
Contributions | Difficulties/challenges | Features/results |
---|---|---|
A multi-abstraction-level knowledge representation approach [5] | Evaluation of the comprehensibility | |
Efficiency of learning; proving the completeness | ||
A measure for the subjectively experienced strategic depth in games [8] | An eligible definition of the measure and its evaluation | A good fit to the strategic depth felt by humans players in [8] when playing GVGAI games |
The GVGAI framework’s strong time constraints (cf. footnote 4) | ||
Practical implementations to be used in further domains [7] | Keeping interfaces lightweight; handling of numeric data | The open-source toolbox InteKRator [13] combining knowledge representation and machine learning aspects |
2 Related Work
-
learning approaches that are able to provide structural insights into what an agent learns
-
approaches that are geared towards a comprehensible representation of knowledge (i.e., that is not only compact but also easy to read and understand for humans)
-
learning/hybrid agent models in the context of games
-
systems similar to the InteKRator toolbox.