Presentation
When Forgetting is Learning: Human-Inspired Memory Management in a Policy Capturing AI System
SessionPoster Session 1
DescriptionArtificial intelligence (AI) systems must adapt quickly to dynamic and interactive environments, but maintaining relevance over time remains a challenge. Inspired by the adaptive advantages of human forgetting, this study investigates the integration of a forgetting function into an AI system. We implemented this mechanism as a training window within the Cognitive Shadow (CS) system, an AI designed to learn and emulate human decision models for classification tasks. This training window hyperparameter – applicable to any supervised machine learning algorithm – helps address the issue of concept drift by prioritizing recent information. The effectiveness of this addition was tested with a simple strategy game similar in dynamics to rock-paper-scissors. Each participant played three sessions of 12 battles, with each consisting of five rounds, against an AI opponent. CS was trained during Session 1 and became active in Sessions 2 and 3. Analyses showed that including the training window significantly improved predictive accuracy of CS in both Sessions 2 and 3 by emphasizing recent, relevant data and filtering out noise. These findings highlight the potential of incorporating human-inspired forgetting mechanisms to enhance AI performance in interactive and dynamic environments, with implications for real-time decision-making support systems and interactive applications.
Event Type
Poster
TimeTuesday, October 14th5:30pm - 6:30pm CDT
LocationRiverside East