Sparse Fourier Transforms on Rank-1 Lattices for the Rapid and Low-Memory Approximation of Functions of Many Variables

12/17/2020
โˆ™
by   Craig Gross, et al.
โˆ™
0
โˆ™

We consider fast, provably accurate algorithms for approximating functions on the d-dimensional torus, f: ๐•‹ ^d โ†’โ„‚, that are sparse (or compressible) in the Fourier basis. In particular, suppose that the Fourier coefficients of f, {c_ k (f) }_ kโˆˆโ„ค^d, are concentrated in a finite set I โŠ‚โ„ค^d so that min_ฮฉโŠ‚ I s.t. |ฮฉ| =s f - โˆ‘_ kโˆˆฮฉ c_ k (f) e^ -2 ฯ€ i kยทโˆ˜_2 < ฯตf _2 holds for s โ‰ช |I| and ฯตโˆˆ (0,1). We aim to identify a near-minimizing subset ฮฉโŠ‚ I and accurately approximate the associated Fourier coefficients { c_ k (f) }_ kโˆˆฮฉ as rapidly as possible. We present both deterministic as well as randomized algorithms using O(s^2 d log^c (|I|))-time/memory and O(s d log^c (|I|))-time/memory, respectively. Most crucially, all of the methods proposed herein achieve these runtimes while satisfying theoretical best s-term approximation guarantees which guarantee their numerical accuracy and robustness to noise for general functions. These are achieved by modifying several one-dimensional Sparse Fourier Transform (SFT) methods to subsample a function along a reconstructing rank-1 lattice for the given frequency set I to rapidly identify a near-minimizing subset ฮฉโŠ‚ I without using anything about the lattice beyond its generating vector. This requires new fast and low-memory frequency identification techniques capable of rapidly recovering vector-valued frequencies in โ„ค^d as opposed to simple integer frequencies in the univariate setting. Two different strategies are proposed and analyzed, each with different accuracy versus computational speed and memory tradeoffs.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset