Data Mining for Hierarchical Model Creation
Authors:
- G. Michael YoungbloodGYG. Michael Youngblood
- Diane J. CookDCDiane J. Cook
- 25 June 2007
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Systems, Man and Cybernetics, Part C (Applications and Reviews)
- Vol. 37 (4) , 561-572
Abstract
In this paper, we examine the problem of learning inhabitant behavioral models in intelligent environments. We maintain that inhabitant interactions in smart environments can be automated using a data-driven approach to generate hierarchical inhabitant models and learn decision policies. To validate this hypothesis, we have designed the ProPHeT decision-learning algorithm that learns a strategy for controlling a smart environment based on sensor observation, power line control, and the generated hierarchical model. The performance of the algorithm is evaluated using real data collected from our MavHome smart home and smart office environments.Keywords
This publication has 17 references indexed in Scilit:
- Online Sequential Prediction via Incremental Parsing: The Active LeZi AlgorithmIEEE Intelligent Systems, 2007
- Motion Detection Using Wavelet Analysis and Hierarchical Markov Models2006
- Managing Adaptive Versatile environmentsPervasive and Mobile Computing, 2005
- Creating an Ambient-Intelligence Environment Using Embedded AgentsIEEE Intelligent Systems, 2004
- Hierarchical recognition of daily human actions based on continuous Hidden Markov Models2004
- Approximate planning with hierarchical partially observable Markov decision process models for robot navigation2003
- Self-segmentation of sequences2003
- The Hierarchical Hidden Markov Model: Analysis and ApplicationsMachine Learning, 1998
- An introduction to hidden Markov modelsIEEE ASSP Magazine, 1986
- Randomness and Mathematical ProofScientific American, 1975