On the Consistency of the Bootstrap Approach for Support Vector Machines and Related Kernel Based Methods

This provides a theoretical foundation for using bootstrap methods to estimate uncertainty in SVMs, which is incremental as it extends consistency results to a broader class of kernel-based methods.

The paper tackled the problem of approximating the finite sample distribution of support vector machines (SVMs) and related kernel methods, showing that bootstrap approximations are consistent for SVMs based on convex and smooth loss functions and general kernels.