Design and Scheduling of an AI-based Queueing System
Researchers have developed an AI-based queueing system designed to optimize scheduling decisions by predicting job classes and accounting for prediction errors. The system's theoretical framework guides the design of predictive models, prioritizing downstream queueing performance. This approach was demonstrated on a content moderation task using toxicity classifiers fine-tuned from large language models. AI
IMPACT Provides a framework for designing AI-powered systems that optimize resource allocation and scheduling in complex service environments.