Deep Learning for Reliable Soft Sensor Development

Soft sensors are widely used in industrial systems to monitor key variables that are difficult to measure, using measurements from other available physical sensors. This thesis aims to develop efficient modeling strategies using deep learning methods to construct data-driven soft sensors that are both accurate and reliable. The outcomes of this research have significant impacts towards advancing the state-of-the-art of deep learning-based soft sensors by enhancing their practicality in modern industries.