Deep and Shallow Architecture of Multilayer Neural Networks

被引：59

作者：

Chang, Chih-Hung ^{[1
]}

机构：

[1] Natl Univ Kaohsiung, Dept Appl Math, Kaohsiung 81148, Taiwan

来源：

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS | 2015年 / 26卷 / 10期

关键词：

Deep architecture; factor-like matrix; multilayer neural networks (MNNs); sofic shift; topological entropy; PATTERN-FORMATION; ALGORITHM;

D O I：

10.1109/TNNLS.2014.2387439

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This paper focuses on the deep and shallow architecture of multilayer neural networks (MNNs). The demonstration of whether or not an MNN can be replaced by another MNN with fewer layers is equivalent to studying the topological conjugacy of its hidden layers. This paper provides a systematic methodology to indicate when two hidden spaces are topologically conjugated. Furthermore, some criteria are presented for some specific cases.

引用

页码：2477 / 2486

页数：10

共 50 条

[41] Compact yet efficient hardware architecture for multilayer-perceptron neural networks
Da Silva R.M.
Nedjah N.
De Macedo Mourelle L.
Controle y Automacao, 2011, 22 (06): : 647 - 663
[42] On the Importance of Network Architecture in Training Very Deep Neural Networks
Chi, Zhizhen
Li, Hongyang
Wang, Jingjing
Lu, Huchuan
2016 IEEE INTERNATIONAL CONFERENCE ON SIGNAL PROCESSING, COMMUNICATIONS AND COMPUTING (ICSPCC), 2016,
[43] Automatic Generation of Dynamic Inference Architecture for Deep Neural Networks
Zhao, Shize
He, Liulu
Xie, Xiaoru
Lin, Jun
Wang, Zhongfeng
2021 IEEE WORKSHOP ON SIGNAL PROCESSING SYSTEMS (SIPS 2021), 2021, : 117 - 122
[44] Architecture-Preserving Provable Repair of Deep Neural Networks
Tao, Zhe
Nawas, Stephanie
Mitchell, Jacqueline
Thakur, Aditya V.
PROCEEDINGS OF THE ACM ON PROGRAMMING LANGUAGES-PACMPL, 2023, 7 (PLDI):
[45] Developing a Hybrid Network Architecture for Deep Convolutional Neural Networks
Sayan, H. Huseyin
Tekgozoglu, O. Faruk
Sonmez, Yusuf
Turan, Bilal
ARTIFICIAL INTELLIGENCE AND APPLIED MATHEMATICS IN ENGINEERING PROBLEMS, 2020, 43 : 750 - 757
[46] DeepRecon: Dynamically Reconfigurable Architecture for Accelerating Deep Neural Networks
Rzayev, Tayyar
Moradi, Saber
Albonesi, David H.
Manohar, Rajit
2017 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2017, : 116 - 124
[47] Daily Prediction Model of Photovoltaic Power Generation Using a Hybrid Architecture of Recurrent Neural Networks and Shallow Neural Networks
Castillo-Rojas, Wilson
Bekios-Calfa, Juan
Hernandez, Cesar
INTERNATIONAL JOURNAL OF PHOTOENERGY, 2023, 2023
[48] An integrated architecture for shallow and deep processing
Crysmann, B
Frank, A
Kiefer, B
Müller, S
Neumann, G
Piskorski, J
Schäfer, U
Siegel, M
Uszkoreit, H
Xu, FY
Becker, M
Krieger, HU
40TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, PROCEEDINGS OF THE CONFERENCE, 2002, : 441 - 448
[49] USING MULTILAYER PERCEPTRON AND DEEP NEURAL NETWORKS FOR THE DIAGNOSIS OF BREAST CANCER CLASSIFICATION
Igodan, C. Efosa
Ukaoha, Kingsley C.
2019 IEEE AFRICON, 2019,
[50] A General-Purpose Neural Architecture Search Algorithm for Building Deep Neural Networks
Zito, Francesco
Cutello, Vincenzo
Pavone, Mario
METAHEURISTICS, MIC 2024, PT II, 2024, 14754 : 126 - 141

← 1 2 3 4 5 →