IKA: Independent Kernel Approximator
This is an incremental improvement for machine learning practitioners working with kernel methods, offering a more flexible approximation approach.
The paper tackles the problem of low rank kernel approximation by introducing IKA, a method that constructs approximations as linear combinations of arbitrarily chosen functions, and it consistently outperformed the Nyström method on the STL-10 dataset.
This paper describes a new method for low rank kernel approximation called IKA. The main advantage of IKA is that it produces a function $ψ(x)$ defined as a linear combination of arbitrarily chosen functions. In contrast the approximation produced by Nyström method is a linear combination of kernel evaluations. The proposed method consistently outperformed Nyström method in a comparison on the STL-10 dataset. Numerical results are reproducible using the source code available at https://gitlab.com/matteo-ronchetti/IKA