Learning group penalties for structured sparsity

The LLSS package contains Matlab code to learn relevances of groups of variables (also referred to as group penalties, or group weights) in the context of structured sparse linear regression. For more details please consult the README.

Graph kernels

During my PhD I wrote some code for computing various kernels for graph-structured data. This zip (560 KB) archive contains Matlab scripts to compute graph kernels for graphs with unlabeled or categorically labeled nodes, such as the random walk, shortest path, graphlet, several instances of Weisfeiler-Lehman or other subtree kernels. For a more detailed list of available kernels please consult the README in the archive.

Note that the Matlab code is not maintained any more. More recent implementations of graph kernels in R and Python, by members of the Machine Learning and Computational Biology Lab at ETH Zürich, are available via GitHub (R package, Python package).

Graph classification data sets

You can also download the graph data sets MUTAG, PTC, NCI1, NCI109, ENZYMES, and D&D (all in Matlab format) as a zip (10 MB) archive. This archive also contains a README describing the data sets and explaining their usage.