PRIMORIS      Contacts      FAQs      INSTICC Portal
 

Tutorials

The role of the tutorials is to provide a platform for a more intensive scientific exchange amongst researchers interested in a particular topic and as a meeting point for the community. Tutorials complement the depth-oriented technical sessions by providing participants with broad overviews of emerging fields. A tutorial can be scheduled for 1.5 or 3 hours.


Dr. Anna Förster
University of Applied Sciences of Southern Switzerland
Switzerland


Machine Learning for Large Scale Sensor Data

Abstract
Machine learning is a family of algorithms and techniques, extremely useful for processing large amounts of fuzzy and unstructured data to learn the underlying model. Especially with the current fast development of cyber-physical systems, such as sensor networks, smart phones and the internet of things, these algorithms will prove to be one of the best approaches for analysis and classification of the resulting large scale data. Furthermore, machine learning addresses also resource-efficient gathering and storage problems for the ever increasing amount of data. In this tutorial we will explore some of the main machine learning techniques, including reinforcement learning, swarm intelligence, decision tree learning, neural networks and some others. We will compare their properties and will evaluate their advantages in the context of various application scenarios and problems. We will see some real world examples for their usage in the general area of cyber-physical systems and especially in the context of big data.

Keywords
machine learning, artificial intelligence, wireless sensor networks, cyber-physical systems, big data, large scale data, sensory data

Duration
3 hours

Aims and Learning Objectives
The main objective of this tutorial is to give researchers the necessary instruments to process, analyze and leverage large scale data from sensor networks. The learnt techniques and approaches can be directly applied to various levels of sensor networking - from the networking stack to high level distributed applications

Biography of Dr. Anna Förster
Dr. Anna Förster is currently a researcher at the Networking Laboratory of the University of Applied Sciences of Southern Switzerland. Dr. Förster has received her Master degree in Computer Science and Aerospace Engineering from the Free University of Berlin, Germany, in 2004 and her Phd in Informatics from the University of Lugano, Switzerland in 2009. In her Phd thesis, she has explored the usage of reinforcement learning to routing and data dissemination in very large wireless sensor networks. She is a co-author of an influential survey paper on computational intelligence for sensor networks and has applied various machine learning techniques to a broad variety of applications and data processing.

Related Experience
Dr. Anna Förster has held some relevant tutorials at:
  • EMERGING conference in Malta, 2009, approx. 40 attendees
  • CONET summer school "Networked Embedded Systems: Humans in the Loop" in Bertinoro, Italy, 2011, approx. 60 attendees
  • European Conference on Wireless Sensor Networks, Bonn, Germany, 2011, approx. 30 attendees The currently proposed tutorial is an extended version, covering especially large data issues and applications.
For example, it will cover techniques for optimal sensing and storage of big data, and data analysis in real world applications based on sensor networks, smart phones and the internet of things.

Contacts
e-mail: sensornets.secretariat@insticc.org

footer