A Geometric Algorithm for Scalable Multiple Kernel Learning
This work addresses the scalability problem for researchers and practitioners in machine learning dealing with large datasets in kernel-based methods, though it is incremental as it builds on existing MKL frameworks.
The paper tackles the scalability issue in Multiple Kernel Learning by reformulating it as a geometric problem to maximize the minimum kernel distance between convex polytopes, resulting in a method that is significantly faster and scales efficiently to larger datasets, with empirical evaluation on eleven datasets showing favorable performance compared to uniform kernel combinations.
We present a geometric formulation of the Multiple Kernel Learning (MKL) problem. To do so, we reinterpret the problem of learning kernel weights as searching for a kernel that maximizes the minimum (kernel) distance between two convex polytopes. This interpretation combined with novel structural insights from our geometric formulation allows us to reduce the MKL problem to a simple optimization routine that yields provable convergence as well as quality guarantees. As a result our method scales efficiently to much larger data sets than most prior methods can handle. Empirical evaluation on eleven datasets shows that we are significantly faster and even compare favorably with a uniform unweighted combination of kernels.