Abstract:The random Fourier features (RFFs) method is a powerful and popular technique in kernel approximation for scalability of kernel methods. The theoretical foundation of RFFs is based on the Bochner theorem (Bochner in Harmonic Analysis and the Theory of Probability, University of California Press, 1995) that relates symmetric, positive definite (PD) functions to probability measures. This condition naturally excludes asymmetric functions with a wide range applications in practice, e.g., directed graphs, conditional probability, and asymmetric kernels. Nevertheless, understanding asymmetric functions (kernels) and its scalability via RFFs is unclear both theoretically and empirically. In this paper, we introduce a complex measure with the real and imaginary parts corresponding to four finite positive measures, which expands the application scope of the Bochner theorem. By doing so, this framework allows for handling classical symmetric, PD kernels via one positive measure; symmetric, non-positive definite kernels via signed measures; and asymmetric kernels via complex measures, thereby unifying them into a general framework by RFFs, named AsK-RFFs. Such approximation scheme via complex measures enjoys theoretical guarantees in the perspective of uniform convergence. In algorithmic implementation, to speed up the kernel approximation process, which is expensive due to the calculation of total masses, we propose a subset-based fast estimation method. This method focuses on optimizing total masses within a sub-training set, effectively transforming the numerical integration for total masses into quadratic programming in three-dimension with low time complexity. AsK-RFFs provides two explicit feature mappings to approximate an asymmetric kernel. These mappings can be utilized in classifiers within the framework of AsK-LS (He et al. in IEEE Trans Pattern Anal Machine Intell 45(8):10044–10054), fulfilling the purpose of using asymmetric kernels in machine learning applications. Our AsK-RFFs method is empirically validated on several typical large-scale datasets and achieves promising kernel approximation performance, which demonstrates the effectiveness of AsK-RFFs.

End-to-end Kernel Learning via Generative Random Fourier Features

Deep Kernel Learning via Random Fourier Features

Self-representative kernel concept factorization

Learning Random Fourier Features by Hybrid Constrained Optimization

RFFNet: Large-Scale Interpretable Kernel Methods via Random Fourier Features

Random Fourier Feature Kernel Recursive Least Squares.

A Double-Variational Bayesian Framework in Random Fourier Features for Indefinite Kernels

Adaptive Random Fourier Features Kernel LMS

Random Fourier Features Multi-Kernel LMS Algorithm

Learning Adaptive Random Features

Random Fourier Features Based Nonlinear Recurrent Kernel Normalized LMS Algorithm with Multiple Feedbacks

Random Fourier Features for Operator-Valued Kernels

Random fourier features for asymmetric kernels

Structure-Aware Random Fourier Kernel for Graphs

Adaptive Random Fourier Features Gaussian Kernel Normalized LMS Algorithm

Automated Spectral Kernel Learning

Stein Random Feature Regression

Random Fourier Features Via Fast Surrogate Leverage Weighted Sampling

Kernel-target Alignment Based Non-Linear Metric Learning

Towards Unbiased Random Features with Lower Variance for Stationary Indefinite Kernels

Learning to Learn Kernels with Variational Random Features