A Method for Sensor Reduction in a Supervised Machine Learning Classification System

Abstract

Smart devices employing interconnected sensors for feedback and control are being rapidly adopted. Many useful applications for these devices are in markets that demand cost-conscious solutions. Traditional machine-learning-based control systems often rely on multiple measurements from many sensors to achieve performance targets. An alternative method is presented that leverages a time-series output produced by a single sensor. By using domain expert knowledge, the time-series output is discretized into finite intervals that correspond to the physical events occurring in the system. Statistical measures are taken across these intervals to serve as the features to the machine learning system. Additional features that decouple key physical metrics are identified, improving the performance of the system. This novel approach requires a more modest dataset and does not compromise performance. The resulting development effort is significantly more cost-effective than traditional sensor classification systems, not only due to the reduced sensor count, but also due to a significantly simplified and more robust algorithm development and testing step. Results are presented with the case study of a media-type classification system within a printing system, which was deployed to the field as a commercial product.