Kernel computations from large-scale random features obtained by Optical Processing Units - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2020

Kernel computations from large-scale random features obtained by Optical Processing Units

Ruben Ohana
  • Fonction : Auteur
Jonas Wacker
  • Fonction : Auteur
Jonathan Dong
  • Fonction : Auteur
Sébastien Marmin
  • Fonction : Auteur
Florent Krzakala
Maurizio Filippone
  • Fonction : Auteur
  • PersonId : 1021042
Laurent Daudet
  • Fonction : Auteur

Résumé

Approximating kernel functions with random features (RFs)has been a successful application of random projections for nonparametric estimation. However, performing random projections presents computational challenges for large-scale problems. Recently, a new optical hardware called Optical Processing Unit (OPU) has been developed for fast and energy-efficient computation of large-scale RFs in the analog domain. More specifically, the OPU performs the multiplication of input vectors by a large random matrix with complex-valued i.i.d. Gaussian entries, followed by the application of an element-wise squared absolute value operation - this last nonlinearity being intrinsic to the sensing process. In this paper, we show that this operation results in a dot-product kernel that has connections to the polynomial kernel, and we extend this computation to arbitrary powers of the feature map. Experiments demonstrate that the OPU kernel and its RF approximation achieve competitive performance in applications using kernel ridge regression and transfer learning for image classification. Crucially, thanks to the use of the OPU, these results are obtained with time and energy savings.

Dates et versions

hal-02987982 , version 1 (04-11-2020)

Identifiants

Citer

Ruben Ohana, Jonas Wacker, Jonathan Dong, Sébastien Marmin, Florent Krzakala, et al.. Kernel computations from large-scale random features obtained by Optical Processing Units. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 2020, Barcelona, Spain. ⟨hal-02987982⟩
131 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More