Abstract. A major challenge in pervasive computing is to learn activity patterns, such as bathing and cleaning from sensor data. Typical sensor deployments generate sparse datasets...
The paper proposes a new method to perform foreground detection by means of background modeling using the tensor concept. Sometimes, statistical modelling directly on image values...
— As the technology scales into 90nm and below, process-induced variations become more pronounced. In this paper, we propose an efficient stochastic method for analyzing the vol...
The Generative Topographic Mapping (GTM) model was introduced by 7) as a probabilistic re-formulation of the self-organizing map (SOM). It offers a number of advantages compared ...
In discretization of a continuous variable its numerical value range is divided into a few intervals that are used in classification. For example, Na¨ıve Bayes can benefit from...