Аннотация:The work is devoted to the analysis of the quality of the transmitted and reproduced multimedia content. End-user quality assessment is based on subjective perception, but it determines whether a person will use a particular product or service. At the same time, most users of multimedia services are not in their own networks of providers with the ability to provide and monitor the required quality of QoS (for example, IPTV) service, but in networks of other operators where information about the deterioration of the quality of the provided service comes post factum in the form of user complaints OTT services). Until now, there are no uniform methods for assessing the quality of service (QoS) and quality of perception (QoE). The proposed methodology is based on the methods of machine learning and neural networks and allows in an automatic mode to assess the quality of perception based on objective parameters. Methods of objectifying and parametrizing the evaluation are also considered.