arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2309.06367cs.HC
收藏

用评估与强化学习建模认知-情感过程

Modeling Cognitive-Affective Processes with Appraisal and Reinforcement Learning

Jiayi Zhang, Joost Broekens, Jussi Jokinen

备注：
15 pages, 7 figures
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出CPM-RL模型，将评估理论与强化学习结合，用时序差分学习形式化CPM评估检查，以建模任务事件引发的情绪反应，并通过情境短文研究验证其预测多种人类情绪的能力。

摘要：

计算模型可以从信息加工的角度揭示认知与情绪之间的相互作用，从而推动情感科学的发展。我们提出了一种融合强化学习（RL）与评估理论的情绪计算模型，在奖励加工、目标导向任务学习、认知评估和情绪体验之间建立了形式化关系。该模型通过用时序差分学习更新来形式化成分过程模型（CPM）中的评估检查，从而实现了这一目标。我们对新颖性、目标相关性、目标助益性和能动性进行了形式化。该形式化与任务无关，可应用于任何能够表示为马尔可夫决策问题（MDP）并用强化学习求解的任务。我们考察了CPM-RL在多大程度上能够模拟由交互式任务事件引发的情绪反应。我们基于一系列情境短文研究预测多种人类情绪来评估该模型，突出其在增进我们对奖励加工在情感体验中作用的理解方面的潜力。

英文摘要：

Computational models can advance affective science by shedding light onto the interplay between cognition and emotion from an information processing point of view. We propose a computational model of emotion that integrates reinforcement learning (RL) and appraisal theory, establishing a formal relationship between reward processing, goal-directed task learning, cognitive appraisal and emotional experiences. The model achieves this by formalizing evaluative checks from the component process model (CPM) in terms of temporal difference learning updates. We formalized novelty, goal relevance, goal conduciveness, and power. The formalization is task independent and can be applied to any task that can be represented as a Markov decision problem (MDP) and solved using RL. We investigated to what extent CPM-RL enables simulation of emotional responses cased by interactive task events. We evaluate the model by predicting a range of human emotions based on a series of vignette studies, highlighting its potential in improving our understanding of the role of reward processing in affective experiences.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑