Probabilistic graphical models for semi-supervised traffic classification

Traffic classification using machine learning continues to be an active research area. The majority of work in this area uses off-the-shelf machine learning tools and treats them as black-box classifiers. This approach turns all the modelling complexity into a feature selection problem. In this paper, we build a problem-specific solution to the traffic classification problem by designing a custom probabilistic graphical model. Graphical models are a modular framework to design classifiers which incorporate domain-specific knowledge. More specifically, our solution introduces semi-supervised learning which means we learn from both labelled and unlabelled traffic flows. We show that our solution performs competitively compared to previous approaches while using less data and simpler features. Copyright © 2010 ACM.

Assessment of the contribution of surface roughness to airframe noise

The generation of sound by turbulent boundary-layer flow at low Mach number over a rough wall is investigated by applying a theoretical model that describes the scattering of the turbulence near field into sound by roughness elements. Attention is focused on the numerical method to approximately quantify the absolute level of far-field radiated roughness noise. Models for the source statistics are obtained by scaling smooth-wall data by the increased skin friction velocity and boundary-layer thickness for a rough surface. Numerical integration is performed to determine the roughness noise, and it reproduces the spectral characteristics of the available empirical formula and experimental data. Experiments are conducted to measure the radiated sound from two rough plates in an open jet The measured noise spectra of the rough plates are above that of a smooth plate in 1-2.5 kHz frequency and exhibit reasonable agreement with the predicted level. Estimates of the roughness noise for a Boeing 757 sized aircraft wing with idealized levels of surface roughness show that hi the high-frequency region the sound radiated from surface roughness may exceed that from the trailing edge, and higher overall sound pressure levels are observed for the roughness noise. The trailing edge noise is also enhanced by surface roughness somewhat A parametric study indicates that roughness height and roughness density significantly affect the roughness noise with roughness height having the dominant effect The roughness noise directivity varies with different levels of surface roughness. Copyright © 2007 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.