Approximation of functionals on Korobov spaces with Fourier Functional Networks.

Journal: Neural networks : the official journal of the International Neural Network Society
PMID:

Abstract

Learning from functional data with deep neural networks has become increasingly useful, and numerous neural network architectures have been developed to tackle high-dimensional problems raised in practical domains. Despite the impressive practical achievements, theoretical foundations underpinning the ability of neural networks to learn from functional data largely remain unexplored. In this paper, we investigate the approximation capacity of a functional neural network, called Fourier Functional Network, consisting of Fourier neural operators and deep convolutional neural networks with a great reduction in parameters. We establish rates of approximating by Fourier Functional Networks nonlinear continuous functionals defined on Korobov spaces of periodic functions. Finally, our results demonstrate dimension-independent convergence rates, which overcomes the curse of dimension.

Authors

  • Peilin Liu
    College of Food Science and Biology, Hebei University of Science and Technology, Shijiazhuang 050018, China.
  • Yuqing Liu
    School of Data Science, City University of Hong Kong, Kowloon, Hong Kong. Electronic address: yuqinliu6-c@my.cityu.edu.hk.
  • Xiang Zhou
    Department of Sociology, Harvard University, Cambridge, Massachusetts, USA.
  • Ding-Xuan Zhou
    School of Data Science and Department of Mathematics, City University of Hong Kong, Kowloon, Hong Kong. Electronic address: mazhou@cityu.edu.hk.