Different industries use data analytics and the process modelling technique successfully in a variety of ways. These popular intelligent approaches improve the quality and quantity of production. This book focuses on the technique of soft-sensing based on spectral data with multi-source high-dimensional mechanical frequency in order to assess difficult-to-measure process parameters. The book will be of interest to researchers and professors working in data analytics, engineers and technicians who need a modelling method based on small sample data, and Ph D students who need to solve modelling and control challenges in a practical way.