Site Network: Cognitive Systems | SFB/TR 8 Spatial Cognition | Fachbereich 03 | Universität Bremen

Lutz Frommberger

Lutz Frommberger's picture

Personal Information

Title Dr.-Ing.
Function Postdoctoral Researcher
Office 3.53
Phone 64281
Fax 9864 - 281
eMail lutz

Affiliation

Logos

Capacity Lab

 

 

Profession

About me

I am a postdoctoral researcher at the Cognitive Systems Group at the University of Bremen. The aim of my research is to enable machines to show adaptive behavior towards new situations, unexpected events, changed circumstances, and interaction with human or artificial partners - that is, in a large and feature-rich space. In my understanding, the question of how to reuse gained knowledge is central for exhibiting such adaptive behavior, and the capability of abstraction is key to allow for that. Thus, my research is centered around abstraction techniques in adaptive artificial systems. Up to 2010, I was a member of the R3-[Q-Shape] project in the DFG collaborate research center SFB/TR 8 Spatial Cognition, where I am still closely associated with. Since 2011, I am executive manager of the International Lab for Local Capacity Building (Capacity Lab).

News

 

  • I am happy to announce that there will be a second edition of the MLIS workshop in 2013. We will host "Machine Learning for Interactive Systems (MLIS): Bridging the Gap Between Perception, Action and Communication" at IJCAI 2013 in Beijing. I am co-organizing MLIS-2013 with Heriberto Cuayahuitl, Nina Dethlefs, and Martijn van Otterlo.
  • Following up the successful MLIS-2012 workshop, I am co-editing a special issue of the ACM Transactions on Interactive Intelligent Systems on Machine Learning for Multiple Modalities in Interactive Systems and Robots. See the Call for Papers, submission deadline was Feb 28, 2013.

 

Interests

My main research interest is the representation of space in applications. I have been working extensively on spatial abstraction techniques for machine learning, in particular reinforcement learning. My main interest is the application of reinforcement learning in large state spaces and real-world systems such as robots, and I am convinced that the representation of space plays a crucial role there. I am also interested in cognitive robotics in general and abstract robot mapping in particular. Some of my research is concerned with processes in logistics and understanding and interpretation of these processes as well as application of cognitively motivated methods to industrial production. Currently, I am applying my research in the field of ICT4D (ICT for Development).

Ph.D. Project

[Book Cover Lutz Frommberger] Qualitative Spatial Abstraction for Reinforcement Learning (concluded August 2009).

The corresponding book Qualitative Abstraction for Reinforcement Learning is available from Springer. A great gift for all your friends and family members interested in machine learning and the role of spatial representations. Contains color graphics!

Research Projects

 

 

Teaching

Current teaching:

I offer diploma/master theses:

Former teaching:

Events Organized
 
  • 2013, August 4 MLIS-2013, 2nd Workshop on Machine Learning for Interactive Systems at IJCAI 2013, Beijing, China
  • 2012, August 27: MLIS-2012, Workshop on Machine Learning for Interactive Systems at ECAI 2012, Montpellier, France
  • 2012, August 28: AILog-2012, 3rd Workshop on Artificial Intelligence and Logistics at ECAI 2012, Montpellier, France
  • 2011, July 16:  AILog-2011, 2nd Workshop on Artificial Intelligence and Logistics at IJCAI 2011, Barcelona, Spain
  • 2010, August 17:  AILog-2010, Workshop on Artificial Intelligence and Logistics at ECAI 2010, Lisbon, Portugal

 

Publications