The Development of an Industrial Learning Classifier System for Data-Mining in a Steel Hop Strip Mill

Industrial domains seek to maximise profits from existing plant due to the large capital costs and long-term payback times of new plant. Modern data-capture technology is comparatively low cost and produces a data resource. This collected data has the potential to give prediction and diagnosis of plant input (e.g. set up) to output (e.g. product quality) from the information contained. Many bespoke and commercial data-mining tools exist, but the novel Artificial Intelligence (AI) technique of Learning Classifier Systems (LCS) has unique properties that could give commercial advantage if developed for such industrial domains.