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.

An approach to scenario analysis of the sustainability of an industrial sector applied to clothing and textiles in the UK

Companies aiming to be 'sustainability leaders' in their sector and governments wanting to support their ambitions need a means to assess the changes required to make a significant difference in the impact of their whole sector. Previous work on scenario analysis/scenario planning demonstrates extensive developments and applications, but as yet few attempts to integrate the 'triple bottom line' concerns of sustainability into scenario planning exercises. This paper, therefore, presents a methodology for scenario analysis of large change to an entire sector. The approach includes calculation of a 'triple bottom line graphic equaliser' to allow exploration and evaluation of the trade-offs between economic, environmental and social impacts. The methodology is applied to the UK's clothing and textiles sector, and results from the study of the sector are summarised. In reflecting on the specific study, some suggestions are made about future application of a similar methodology, including a template of candidate solutions that may lead to significant reduction in impacts. © 2007 Elsevier Ltd. All rights reserved.