Innovative algorithms for completion of resource intensive tasks in IoT devices and novel applications in the Smart City & Smart Building

Abstract

This thesis is framed on the topic of Machine Learning, where we have been focused on the refinement of different methods from the literature, and diverse applications related to Smart Cities and Edge Computing. Preciselly, the main contributions have been made by improving algorithms to ease their computation in resource constrained devices, establishing policies for orchestrating load distribution between these devices through long periods of time, opening the way to novel applications. Contributions are focused on: (1) Neural Network reduction, (2) Task offloading in Edge Computing and (3) Building prediction in Smart Cities.