Kernel methods are a class of methods for data analysis that generalize existing techniques by implicitly mapping the data into a high dimensional feature space. The most popular case is Support Vector Machines (SVM). The choice of the kernel and its parameters is crucial for the performance of SVM in terms of classification accuracy.
The default approach for tuning the parameters of the kernel in Support Vector Machine (SVM) is a grid search in the parameter space. Different metaheuristics have been proposed as a more efficient alternative, but they have only shown to be useful in models with a low number of parameters. Complex models, involving many parameters, can be seen as extensions of simpler and easy-to-tune models, yielding a nested sequence of models of increasing complexity. We propose an algorithm which successfully exploits this nested property. We discuss extensions of this method to other problems, with a particular focus on kernel clustering in a dynamic context where the groups and the features evolve over time.
Current 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996