论文标题
Rinascimento：使用事件值函数发挥出色
Rinascimento: using event-value functions for playing Splendor
论文作者
论文摘要
在游戏研究领域，人工通用情报算法通常将得分用作学习或演奏动作的主要奖励信号。但是，这表明当点奖励非常罕见或缺乏直到游戏结束时，其严重的局限性。本文提出了一种基于事件记录的新方法：游戏状态每次其功能都会发生变化时触发事件。这些事件由事件值函数（EF）处理，该函数将值分配给单个动作或序列。实验表明，这种方法可以减轻稀缺点奖励的问题并改善AI性能。此外，这代表了通过描述通过EF的更丰富和可控制的行为空间来控制人造代理采用的策略的向前迈出的一步。 Tuned EF能够整洁地综合游戏中事件的相关性。与多个对手玩游戏时，使用EF的代理商表现出更强的表现。
In the realm of games research, Artificial General Intelligence algorithms often use score as main reward signal for learning or playing actions. However this has shown its severe limitations when the point rewards are very rare or absent until the end of the game. This paper proposes a new approach based on event logging: the game state triggers an event every time one of its features changes. These events are processed by an Event-value Function (EF) that assigns a value to a single action or a sequence. The experiments have shown that such approach can mitigate the problem of scarce point rewards and improve the AI performance. Furthermore this represents a step forward in controlling the strategy adopted by the artificial agent, by describing a much richer and controllable behavioural space through the EF. Tuned EF are able to neatly synthesise the relevance of the events in the game. Agents using an EF show more robust when playing games with several opponents.
