Efficient Approximation of Molecular Kinetics using Random Fourier Features

TL;DR

This paper introduces a computationally efficient method using Random Fourier Features to approximate molecular kinetics, enabling accurate analysis of slow processes in large datasets with improved hyper-parameter tuning.

Contribution

It presents a novel RFF-based approach for kernel approximation in molecular kinetics, reducing computational complexity and enhancing model selection via VAMP score.

Findings

Accurate estimation of slow molecular kinetics in benchmark systems

Efficient hyper-parameter tuning using VAMP score

Reduced computational cost compared to traditional kernel methods

Abstract

Slow kinetic processes of molecular systems can be analyzed by computing dominant eigenpairs of the Koopman operator or its generator. In this context, the Variational Approach to Markov Processes (VAMP) provides a rigorous way of discerning the quality of different approximate models. Kernel methods have been shown to provide accurate and robust estimates for slow kinetic processes, but are sensitive to hyper-parameter selection, and require the solution of large-scale generalized eigenvalue problems, which can easily become computationally demanding for large data sizes. In this contribution, we employ a stochastic approximation of the kernel based on random Fourier features (RFFs), to derive a small-scale dual eigenvalue problem which can easily be solved. We provide an interpretation of this procedure in terms of a finite randomly generated basis set. By combining the RFF approach and model selection by means of the VAMP score, we show that kernel parameters can be efficiently tuned, and accurate estimates of slow molecular kinetics can be obtained for several benchmarking systems, such as deca alanine and the NTL9 protein.