Fourier Sparse Leverage Scores and Approximate Kernel Learning
Tamas Erdelyi, Cameron Musco, Christopher Musco
Spotlight presentation: Orals & Spotlights Track 17: Kernel Methods/Optimization
on 2020-12-09T07:20:00-08:00 - 2020-12-09T07:30:00-08:00
on 2020-12-09T07:20:00-08:00 - 2020-12-09T07:30:00-08:00
Poster Session 4 (more posters)
on 2020-12-09T09:00:00-08:00 - 2020-12-09T11:00:00-08:00
GatherTown: Theory ( Town D0 - Spot C1 )
on 2020-12-09T09:00:00-08:00 - 2020-12-09T11:00:00-08:00
GatherTown: Theory ( Town D0 - Spot C1 )
Join GatherTown
Only iff poster is crowded, join Zoom . Authors have to start the Zoom call from their Profile page / Presentation History.
Only iff poster is crowded, join Zoom . Authors have to start the Zoom call from their Profile page / Presentation History.
Toggle Abstract Paper (in Proceedings / .pdf)
Abstract: We prove new explicit upper bounds on the leverage scores of Fourier sparse functions under both the Gaussian and Laplace measures. In particular, we study s-sparse functions of the form $f(x) = \sum_{j=1}^s a_j e^{i \lambda_j x}$ for coefficients $a_j \in C$ and frequencies $\lambda_j \in R$. Bounding Fourier sparse leverage scores under various measures is of pure mathematical interest in approximation theory, and our work extends existing results for the uniform measure [Erd17,CP19a]. Practically, our bounds are motivated by two important applications in machine learning: 1. Kernel Approximation. They yield a new random Fourier features algorithm for approximating Gaussian and Cauchy (rational quadratic) kernel matrices. For low-dimensional data, our method uses a near optimal number of features, and its runtime is polynomial in the *statistical dimension* of the approximated kernel matrix. It is the first ``oblivious sketching method'' with this property for any kernel besides the polynomial kernel, resolving an open question of [AKM+17,AKK+20b]. 2. Active Learning. They can be used as non-uniform sampling distributions for robust active learning when data follows a Gaussian or Laplace distribution. Using the framework of [AKM+19], we provide essentially optimal results for bandlimited and multiband interpolation, and Gaussian process regression. These results generalize existing work that only applies to uniformly distributed data.