Multi-objective Adaptation of a Parameterized GVGAI Agent Towards Several Games

Abstract

This paper proposes a benchmark for multi-objective optimization based on video game playing. The challenge is to optimize an agent to perform well on several different games, where each objective score corresponds to the performance on a different game. The benchmark is inspired from the quest for general intelligence in the form of general game playing, and builds on the General Video Game AI (GV-GAI) framework. As it is based on game-playing, this benchmark incorporates salient aspects of game-playing problems such as discontinuous feedback and a non-trivial amount of stochasticity. We argue that the proposed benchmark thus provides a different challenge from many other benchmarks for multi-objective optimization algorithms currently available. We also provide initial results on categorizing the space offered by this benchmark and applying a standard multi-objective optimization algorithm to it.

General Video Game AI: A Multitrack Framework for Evaluating Agents, Games, and Content Generation Algorithms

“…In the games tested, the rate of victories grew from 32.24% (normal MCTS) to 42.38% in the multi-objective version, showing great promise for this approach. In a different study, Khalifa et al [31] applied multi-objective concepts to evolving parameters for a tree selection confidence bounds equation. A previous work by Bravi [32] (also discussed in Section IV-D) provided multiple UCB equations for different games.…”

“…A previous work by Bravi [32] (also discussed in Section IV-D) provided multiple UCB equations for different games. The work in [31] evolved, using S-Metric Selection Evolutionary Multi-objective Optimization Algorithm (SMS-EMOA), the linear weights of a UCB equation that results of combining all from [32] in a single one. All these components respond to different and conflicting objectives, and their results show that it is possible to find good solutions for the games tested.…”

Analysing Results from AI Benchmarks: Key Indicators and How to Obtain Them

“…If we focus on the GVGAI methods because of its wider variety of algorithms, we see that AIJim, a variant of MCTS that performs well in several related domains (see [Browne et al, 2012]), has high generality. As for the low generality of NovTea, it is an Iterated Widthbased approach [Lipovetzky and Geffner, 2012], originally a planning technique, which tries to outperform MCTS in GVGAI with specific tuning (pruning using novelty test) [Bontrager et al, 2016].…”

“…Ultimately, it is crucial for AI researchers to know whether they are progressing through generality or through the exploitation of specific subfamilies of problems. The analysis under these key indicators represents a novel way of understanding not only benchmark results in AI , but also video game competitions (e.g., Super Mario Bros [Togelius et al, 2013], Angry Birds [Renz et al, 2015] or StarCraft AI competitions) as well as the existing architectures for multi-purpose game agents and bots addressing them [Hosu and Urzica, 2015], [Khalifa et al, 2017] . This kind of assessment may have a huge impact on how players and competitions are designed and how the results of the AI systems (and humans) are interpreted.…”