Section: New Results

Semi-supervised machine learning

Participant : Paulo Gonçalves [correspondant] .

This contribution is part of the PhD work of M. Sokol (EPI MAESTRO, Oct. 2009 – May 2014), co-supervised with K. Avrachenkov and Ph. Nain, on the classification of content and users in peer-to-peer networks using graph-based semi-supervised learning methods. Semi-supervised learning methods constitute a category of machine learning methods which use labelled points together with unlabelled data to tune the classifier. The main idea of the semi-supervised methods is based on an assumption that the classification function should change smoothly over a similarity graph, which represents relations among data points. This idea can be expressed using kernels on graphs such as graph Laplacian. Different semi-supervised learning methods have different kernels which reflect how the underlying similarity graph influences the classification results. In a recent work, we analysed a general family of semi-supervised methods, provided insights about the differences among the methods and gave recommendations for the choice of the kernel parameters and labelled points. In particular, it appeared that it was preferable to choose a kernel based on the properties of the labelled points. We illustrated our general theoretical conclusions with an analytically tractable characteristic example, clustered preferential attachment model and classification of content in P2P networks. (See [8] )