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.

Towards measuring the impact of information availability on maintenance decisions

Decisions concerning maintenance have become increasingly important and requires a diverse set of information as systems become more complex. The availability of information has an impact on the effectiveness of these decisions, and thus on the performance of the asset. This paper highlights the importance of quantifying the value of information on maintenance decisions and asset performance. In particular, we emphasise the need to focus on measuring value as opposed to cost of maintenance, which is the current practice. In this direction, we propose a measure - Value of Ownership (VOO) - to assess the value of information and performance of maintenance decisions throughout an assets lifecycle. © 2009 IFAC.