QoE Estimation in mobile networks using Machine Learning

Quality of experience (QoE) can be defined as the overall level of acceptability of an application or service, as perceived by the end-user.
The perceived QoE of mobile user plays a key role in the business of the telecom carriers. This work has focused in training a model capable of predicting the QoE of the end-user using a number of Machine Learning approaches, based on quality of service (QoS) metrics from different sources like the mobile device, the mobile network and also subjective metrics given by the user (QoE and Mood surveys) in a real life setup. An android app, a metric collection platform, a system for data processing and semi-automatic analysis of metrics has been developed as a part of this work. The experimental results show that by assembling a combined model of the algorithms with best observed individual performance, improvements in the overall performance of the prediction can be achieved.