Learning general sparse additive models from point queries in high dimensions - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles Constructive Approximation Year : 2019

Learning general sparse additive models from point queries in high dimensions

Abstract

We consider the problem of learning a $d$-variate function $f$ defined on the cube $[−1, 1]^d ⊂ R^d$ , where the algorithm is assumed to have black box access to samples of f within this domain. Denote $S_r ⊂ {[d] \choose r} ; r = 1,. .. , r_0$ to be sets consisting of unknown $r$-wise interactions amongst the coordinate variables. We then focus on the setting where f has an additive structure, i.e., it can be represented as $f = \sum_{j∈S1} φ_j + \sum_{j∈S2} φ_j + · · · + \sum_{j∈S_{r_0}} φ_j$, where each $φ_j ; j ∈ S_r$ is at most r-variate for $1 ≤ r ≤ r_0$. We derive randomized algorithms that query f at carefully constructed set of points, and exactly recover each $S_r$ with high probability. In contrary to the previous work, our analysis does not rely on numerical approximation of derivatives by finite order differences.
Fichier principal
Vignette du fichier
SPAMgen_main.pdf (619.44 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02379404 , version 1 (25-11-2019)

Identifiers

Cite

Hemant Tyagi, Jan Vybiral. Learning general sparse additive models from point queries in high dimensions. Constructive Approximation, 2019, 50, pp.403-455. ⟨10.1007/s00365-019-09461-6⟩. ⟨hal-02379404⟩
36 View
134 Download

Altmetric

Share

Gmail Facebook X LinkedIn More