A method has been developed for express-assessment of the quality of dairy products according to the indicators of optical sensors in the visible and near-IR wavelength range. The use of modern machine learning methods, in particular the principal component method, made it possible to identify groups of samples similar in their properties and determine whether products belong to an industrial or piece manufacturing method. The technique allows you to designate a group of ‘references’, deviations from it, and is an inexpensive express method for controlling the quality of food products.