Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After Learning

Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After Learning

One of the current trends in the deep neural network technology consists in allowing a man–machine interaction and providing an explanation of network design and learning principles. In this direction, an experience with fuzzy systems is of great support. We propose our insight that is based on the...

Full description

Bibliographic Details
Main Authors: Vojtech Molek, Irina Perfilieva
Format: Article
Language:English
Published: Atlantis Press 2020-09-01
Series:International Journal of Computational Intelligence Systems
Subjects:
F-transform
Convolutional neural network
Deep learning
Interpretability
Online Access:https://www.atlantis-press.com/article/125944627/view
id doaj-6b32c6ee42004c6e99c4fe5e5a910e18
record_format Article
spelling doaj-6b32c6ee42004c6e99c4fe5e5a910e182020-11-25T03:52:48ZengAtlantis PressInternational Journal of Computational Intelligence Systems 1875-68832020-09-0113110.2991/ijcis.d.200907.001Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After LearningVojtech MolekIrina PerfilievaOne of the current trends in the deep neural network technology consists in allowing a man–machine interaction and providing an explanation of network design and learning principles. In this direction, an experience with fuzzy systems is of great support. We propose our insight that is based on the particular theory of fuzzy (F)-transforms. Besides a theoretical explanation, we develop a new architecture of a deep neural network where the F-transform convolution kernels are used in the first two layers. Based on a series of experiments, we demonstrate the suitability of the F-transform-based deep neural network in the domain of image processing with the focus on recognition. Moreover, we support our insight by revealing the similarity between the F-transform and first-layer kernels in the most used deep neural networks.https://www.atlantis-press.com/article/125944627/viewF-transformConvolutional neural networkDeep learningInterpretability
collection DOAJ
language English
format Article
sources DOAJ
author Vojtech Molek
Irina Perfilieva
spellingShingle Vojtech Molek
Irina Perfilieva
Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After Learning
International Journal of Computational Intelligence Systems
F-transform
Convolutional neural network
Deep learning
Interpretability
author_facet Vojtech Molek
Irina Perfilieva
author_sort Vojtech Molek
title Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After Learning
title_short Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After Learning
title_full Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After Learning
title_fullStr Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After Learning
title_full_unstemmed Deep Learning and Higher Degree F-Transforms: Interpretable Kernels Before and After Learning
title_sort deep learning and higher degree f-transforms: interpretable kernels before and after learning
publisher Atlantis Press
series International Journal of Computational Intelligence Systems
issn 1875-6883
publishDate 2020-09-01
description One of the current trends in the deep neural network technology consists in allowing a man–machine interaction and providing an explanation of network design and learning principles. In this direction, an experience with fuzzy systems is of great support. We propose our insight that is based on the particular theory of fuzzy (F)-transforms. Besides a theoretical explanation, we develop a new architecture of a deep neural network where the F-transform convolution kernels are used in the first two layers. Based on a series of experiments, we demonstrate the suitability of the F-transform-based deep neural network in the domain of image processing with the focus on recognition. Moreover, we support our insight by revealing the similarity between the F-transform and first-layer kernels in the most used deep neural networks.
topic F-transform
Convolutional neural network
Deep learning
Interpretability
url https://www.atlantis-press.com/article/125944627/view
work_keys_str_mv AT vojtechmolek deeplearningandhigherdegreeftransformsinterpretablekernelsbeforeandafterlearning
AT irinaperfilieva deeplearningandhigherdegreeftransformsinterpretablekernelsbeforeandafterlearning
_version_ 1724480886138732544

Similar Items