Reinforcement Learning, specifically Deep Q-Networks, was applied to semiconductor manufacturing data to efficiently identify substandard products, with a model optimized using the F1-Metric achieving 87% accuracy and a 22.4% time-saving, highlighting RL's potential in enhancing manufacturing processes.
This repo supports a research paper which will be handed in soon. Any updates will be made available soon.