Quadrature-based features for kernel approximation
This work addresses a known bottleneck in kernel methods for machine learning practitioners, offering an incremental improvement over existing approximation techniques.
The paper tackles the problem of improving kernel approximation for scaling kernel methods to larger datasets by proposing a quadrature-based approach that unifies and extends previous random feature methods, achieving better estimates with demonstrated convergence behavior and empirical validation.
We consider the problem of improving kernel approximation via randomized feature maps. These maps arise as Monte Carlo approximation to integral representations of kernel functions and scale up kernel methods for larger datasets. Based on an efficient numerical integration technique, we propose a unifying approach that reinterprets the previous random features methods and extends to better estimates of the kernel approximation. We derive the convergence behaviour and conduct an extensive empirical study that supports our hypothesis.