Grstats Forum
AUEB Stats Webinar 22/10/2020:   Scalable Gaussian Processes, with Guarantees: Kernel Approximations and Deep Feature Extraction by P. Dellaportas Forumgrstats
Grstats Forum
Θέλετε να αντιδράσετε στο μήνυμα; Φτιάξτε έναν λογαριασμό και συνδεθείτε για να συνεχίσετε.

AUEB Stats Webinar 22/10/2020: Scalable Gaussian Processes, with Guarantees: Kernel Approximations and Deep Feature Extraction by P. Dellaportas

Πήγαινε κάτω

AUEB Stats Webinar 22/10/2020:   Scalable Gaussian Processes, with Guarantees: Kernel Approximations and Deep Feature Extraction by P. Dellaportas Empty AUEB Stats Webinar 22/10/2020: Scalable Gaussian Processes, with Guarantees: Kernel Approximations and Deep Feature Extraction by P. Dellaportas

Δημοσίευση από grstats Την / Το Πεμ 1 Οκτ 2020 - 12:39

ΚΥΚΛΟΣ ΣΕΜΙΝΑΡΙΩΝ ΣΤΑΤΙΣΤΙΚΗΣ ΟΚΤΩΒΡΙΟΣ 2020

AUEB Stats Webinar 22/10/2020:   Scalable Gaussian Processes, with Guarantees: Kernel Approximations and Deep Feature Extraction by P. Dellaportas 2020-110


Πέτρος Δελλαπόρτας
Professor in Statistical Science, Department of Statistical Science, University College London και Professor of Statistics, Department of Statistics, Athens University of Economics and Business

Scalable Gaussian Processes, with Guarantees: Kernel Approximations and Deep Feature Extraction

ΠΕΜΠΤΗ 22/10/2020
12:30

Σύνδεσμος Google Meeting: meet.google.com/iut-yhkj-uqh

ΠΕΡΙΛΗΨΗ

We provide a linear time inferential framework for Gaussian processes that supports automatic feature extraction through deep neural networks and low-rank kernel approximations. Importantly, we derive approximation guarantees bounding the Kullback-Leibler divergence between the idealized Gaussian process and one resulting from a low-rank approximation to its kernel under two types of approximations, which result in two instantiations of our framework: Deep Fourier Gaussian Processes, resulting from random Fourier feature low-rank approximations, and Deep Mercer Gaussian Processes, resulting from truncating the Mercer expansion of the kernel. We do extensive experimental evaluation of these two instantiations in a broad collection of real-world datasets providing strong evidence that they outperform a broad range of state-of-the-art methods in terms of time efficiency, negative log-predictive density, and root mean squared error.

Facebook event: https://www.facebook.com/events/1028215017627341
AUEB web link: https://www.dept.aueb.gr/el/stat/events/dellaportas22_10
grstats
grstats

Posts : 819
Join date : 21/10/2009

http://stat-athens.aueb.gr/~grstats/

Επιστροφή στην κορυφή Πήγαινε κάτω

Επιστροφή στην κορυφή


 
Δικαιώματα σας στην κατηγορία αυτή
Δεν μπορείτε να απαντήσετε στα Θέματα αυτής της Δ.Συζήτησης