[Back]

F. Li, S. Dustdar:
"Incorporating Unsupervised Learning in Activity Recognition";
Talk: Activity Context Representation: Techniques and Languages at AAAI-11, San Francisco, California, USA; 2011-08-07 - 2011-08-11; in: "Activity Context Representation: Techniques and Languages at AAAI-11 Technical Report WS-11-04", AAAI Press, WS-11-04 (2011), ISBN: 978-1-57735-520-5; 38 - 41.

Users are constantly involved in a multitude of activities
in ever-changing context. Analyzing activities in contextrich
environments has become a great challenge in contextawareness
research. Traditional methods for activity recognition,
such as classification, cannot cope with the variety and
dynamicity of context and activities. In this paper, we propose
an activity recognition approach that incorporates unsupervised
learning.We analyze the feasibility of applying subspace
clustering-a specific type of unsupervised learning-
to high-dimensional, heterogeneous sensory input. Then we
present the correspondence between clustering output and
classification input. This approach has the potential to discover
implicit, evolving activities, and can provide valuable
assistance to traditional classification based methods.

Project Head Schahram Dustdar:
Cloud computing research lab

Project Head Schahram Dustdar:
SM4ALL