Machine learning based Quality of Experience (QoE) Prediction Approach in Enterprise Multimedia Networks

Abstract

In this paper, we discuss quality of experience in multimedia networks. We present an architecture and a survey of machine learning methods to predict the quality of experience in an enterprise multimedia network environment. Our approach is based on subjective methods. It consists of the use PRTG (Paessler Router Traffic Grapher) for QoS (quality of service) data collection and Google Forms for the different users of the network MOS (Minimum Score Opinion) parameters collection. We then implement different supervised machine learning schemes using the data collected, and finally analyze their performance. We compare two classes of algorithms namely regression algorithms and classification algorithms. The Random Forest Classifier in the second class algorithm give the best results.