报告人简介:

Dr. S V N Vishwanathan a senior researcher in the Statistical Machine Learning program, National ICT Australia with an adjunct appointment at the Research School for Information Sciences and Engineering (RSISE), Australian National University. He research on Machine Learning mainly involves a fusion between ideas from computer science, mathematics and optimization.

报告简介：

We present a unified framework to study graph kernels, special cases of which include the random walk graph kernel, marginalized graph kernel, and geometric kernel on graphs. Through extensions of linear algebra to Reproducing Kernel Hilbert Spaces (RKHS) and reduction to a Sylvester equation, we construct an algorithm that improves the time complexity of our kernel computations from $O(n^6)$ to $O(n^3)$.  When the graphs are sparse, conjugate gradient solvers or fixed-point iterations bring our algorithm into the sub-cubic domain. Experiments on graphs from bioinformatics and other application domains show that it is often more than a thousand times faster than previous approaches.

Time permitting we will also explore connections between diffusion kernels, regularization on graphs, and graph kernels, and use these connections to propose new graph kernels. We will also show that rational kernels when specialized to graphs reduces to the random walk graph kernel, and that the optimal assignment kernel is not a valid positive semi-definite kernel.