Optimality Implies Kernel Sum Classifiers are Statistically Efficient
This work provides theoretical justification for assumptions in multiple kernel learning, but it is incremental as it builds on existing optimization and learning theory tools.
The authors tackled the problem of analyzing the sample complexity of optimal kernel sum classifiers by combining optimization tools with learning theory bounds, showing that these classifiers are statistically efficient, which contrasts with typical results for all classifiers and justifies prior assumptions in multiple kernel learning.
We propose a novel combination of optimization tools with learning theory bounds in order to analyze the sample complexity of optimal kernel sum classifiers. This contrasts the typical learning theoretic results which hold for all (potentially suboptimal) classifiers. Our work also justifies assumptions made in prior work on multiple kernel learning. As a byproduct of our analysis, we also provide a new form of Rademacher complexity for hypothesis classes containing only optimal classifiers.