Skip to Main Content
In the past years, mobile devices were limited to textual content. However, the current generation has started to access richer multimedia content such as video, increasing the diversity of devices accessing the Web. Then, a problem arises as some of those devices characteristics like memory capacity or screen resolution turn the access to a content restricted. The present work considers the use of machine learning techniques as part of a dynamic video adaptation process, comparing the results from two of the most used approaches for data analysis, Multilayer Perceptron and Bayesian Inference, as part of a Decision Engine, analyzing data like device's capabilities, user's preferences and network condition in order to take the most appropriate way to adapt a video stream.