Learning with Dynamic Architectures for Artificial Neural Networks - Adaptive Batch Size Approach

被引：0

作者：

Saeed, Reham ^{[1
]}

Ghnemat, Rawan ^{[1
]}

Benbrahim, Ghassen ^{[2
]}

Elhassan, Ammar ^{[1
]}

机构：

[1] Princess Sumaya Univ Technol, King Hussein Sch Comp Sci, Amman, Jordan

[2] Prince Mohammad bin Fahd Univ, Coll Comp Engn & Sci, Khobar, Saudi Arabia

来源：

2019 2ND INTERNATIONAL CONFERENCE ON NEW TRENDS IN COMPUTING SCIENCES (ICTCS) | 2019年

关键词：

Artificial Neural Networks; Convolutional Neural Networks; Batch Size; Convergence; Accuracy; Training; Feed-forward; Two-Layer Feed-Forward Net; Sampling; Tensorflow; ADANet; Stochasticity; DESIGN;

D O I：

10.1109/ictcs.2019.8923070

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

In this research we explore the performance of ADANET framework by using custom search space for an image-classification dataset using tensorflow libraries in combination with adaptive batch sizes for learning. In one experiment we classified fashion MNISET data and MNIST data of handwritten digits and obtained favorable results in terms of training time as well as accuracy by alternating learning batch sizes dynamically. Our testing was applied using simple deep neural network (DNN) and also with convolutional neural network (CNN).

引用

页码：302 / 305

页数：4

共 50 条

[21] Adaptive post-linearization of dynamic nonlinear systems with artificial neural networks
Rabinowitz, M
Gutt, GM
Franklin, GF
JOURNAL OF DYNAMIC SYSTEMS MEASUREMENT AND CONTROL-TRANSACTIONS OF THE ASME, 1999, 121 (04): : 678 - 685
[22] Adapting the Size of Artificial Neural Networks Using Dynamic Auto-Sizing
Cahlik, Vojtech
Kordik, Pavel
Cepek, Miroslav
2022 IEEE 17TH INTERNATIONAL CONFERENCE ON COMPUTER SCIENCES AND INFORMATION TECHNOLOGIES (CSIT), 2022, : 592 - 596
[23] Learning improvement of spiking neural networks with dynamic adaptive hyperparameter neurons
Liang, Jiakai
Wang, Chao
Ma, De
Li, Ruixue
Yue, Keqiang
Li, Wenjun
APPLIED INTELLIGENCE, 2024, 54 (19) : 9158 - 9176
[24] Effect of batch learning in multilayer neural networks
Fukumizu, K
ICONIP'98: THE FIFTH INTERNATIONAL CONFERENCE ON NEURAL INFORMATION PROCESSING JOINTLY WITH JNNS'98: THE 1998 ANNUAL CONFERENCE OF THE JAPANESE NEURAL NETWORK SOCIETY - PROCEEDINGS, VOLS 1-3, 1998, : 67 - 70
[25] Building Adaptive Tutoring Model using Artificial Neural Networks and Reinforcement Learning
Fenza, Giuseppe
Orciuoli, Francesco
Sampson, Demetrios G.
2017 IEEE 17TH INTERNATIONAL CONFERENCE ON ADVANCED LEARNING TECHNOLOGIES (ICALT), 2017, : 460 - 462
[26] Trajectory tracking control by an adaptive iterative learning control with artificial neural networks
Yamakita, M
Ueno, M
Sadahiro, T
PROCEEDINGS OF THE 2001 AMERICAN CONTROL CONFERENCE, VOLS 1-6, 2001, : 1253 - 1255
[27] A dynamic architecture for artificial neural networks
Ghiassi, M
Saidane, H
NEUROCOMPUTING, 2005, 63 : 397 - 413
[28] A New Probabilistic Approach to On-Line Learning in Artificial Neural Networks
Jankovic, Marko V.
Rubens, Neil
PROCEEDINGS OF THE 3RD INT CONF ON APPLIED MATHEMATICS, CIRCUITS, SYSTEMS, AND SIGNALS/PROCEEDINGS OF THE 3RD INT CONF ON CIRCUITS, SYSTEMS AND SIGNALS, 2009, : 226 - +
[29] Identification with dynamic neural networks - Architectures, comparisons, applications
Isermann, R
Ernst, S
Nelles, O
(SYSID'97): SYSTEM IDENTIFICATION, VOLS 1-3, 1998, : 947 - 972
[30] Simulated annealing based learning approach for the design of cascade architectures of fuzzy neural networks
Han, Chang-Wook
Park, Jung-Il
ADVANCES IN NEURAL NETWORKS - ISNN 2006, PT 1, 2006, 3971 : 798 - 803

← 1 2 3 4 5 →