Παρακαλώ χρησιμοποιήστε αυτό το αναγνωριστικό για να παραπέμψετε ή να δημιουργήσετε σύνδεσμο προς αυτό το τεκμήριο:
https://hdl.handle.net/20.500.14279/19071
Τίτλος: | Nonparametric Bayesian deep networks with local competition | Συγγραφείς: | Panousis, Konstantinos P. Chatzis, Sotirios P. Theodoridis, Sergios |
Major Field of Science: | Natural Sciences | Field Category: | Computer and Information Sciences | Λέξεις-κλειδιά: | Inference engines;Machine learning;Bayesian inference;Benchmark datasets;Model complexity;Network parameters;Non-parametric Bayesian;Predictive accuracy;State of the art;Technical innovation | Ημερομηνία Έκδοσης: | Ιου-2019 | Πηγή: | 36th International Conference on Machine Learning, 2019, 9-15 June, Long Beach, United States | Conference: | International Conference on Machine Learning | Περίληψη: | The aim of this work is to enable inference of deep networks that retain high accuracy for the least possible model complexity, with the latter deduced from the data during inference. To this end, we revisit deep networks that comprise competing linear units, as opposed to nonlinear units that do not entail any form of (local) competition. In this context, our main technical innovation consists in an inferential setup that leverages solid arguments from Bayesian nonparamctrics. Wc infer both the needed set of connections or locally competing sets of units, as well as the required floatingpoint precision for storing the network parameters. Specifically, we introduce auxiliary discrete latent variables representing which initial network components are actually needed for modeling the data at hand, and perform Bayesian inference over them by imposing appropriate stick-breaking priors. As we experimentally show using benchmark datasets, our approach yields networks with less computational footprint than the state-of-the-art, and with no compromises in predictive accuracy. | URI: | https://hdl.handle.net/20.500.14279/19071 | Rights: | Copyright 2019 by the author(s). | Type: | Conference Papers | Affiliation: | National and Kapodistrian University of Athens Cyprus University of Technology |
Publication Type: | Peer Reviewed |
Εμφανίζεται στις συλλογές: | Δημοσιεύσεις σε συνέδρια /Conference papers or poster or presentation |
Αρχεία σε αυτό το τεκμήριο:
Αρχείο | Περιγραφή | Μέγεθος | Μορφότυπος | |
---|---|---|---|---|
1805.07624.pdf | Fulltext | 558.03 kB | Adobe PDF | Δείτε/ Ανοίξτε |
CORE Recommender
Page view(s) 50
340
Last Week
0
0
Last month
4
4
checked on 4 Οκτ 2024
Download(s) 50
81
checked on 4 Οκτ 2024
Google ScholarTM
Check
Αυτό το τεκμήριο προστατεύεται από άδεια Άδεια Creative Commons