Image Classification Based on Automatic Neural Architecture Search Using Binary Crow Search Algorithm

Neural architectures have accelerated the advancement in various domains by enabling automatic pattern detection, image classification, audio recognition, and face recognition etc. However, they are computationally expensive to design and expert knowledge in various domains is required. In this pape...

Full description

Saved in:
Bibliographic Details
Published in:IEEE access Vol. 8; pp. 189891 - 189912
Main Authors: Ahmad, Mobeen, Abdullah, Muhammad, Moon, Hyeonjoon, Yoo, Seong Joon, Han, Dongil
Format: Journal Article
Language:English
Published: Piscataway IEEE 2020
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Algorithms
AutoML
Computer architecture
crow search algorithm
deep learning
Domains
Encoding
Error analysis
Face recognition
Genetic algorithms
Heuristic algorithms
Heuristic methods
hyperparameter optimization
Image classification
Machine learning
metaheuristic
Neural architecture search
Neural networks
Object recognition
Optimization
Parameters
Search algorithms
Search problems
Space exploration
Statistical analysis
Swarm intelligence
ISSN:2169-3536, 2169-3536
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Neural architectures have accelerated the advancement in various domains by enabling automatic pattern detection, image classification, audio recognition, and face recognition etc. However, they are computationally expensive to design and expert knowledge in various domains is required. In this paper, a swarm intelligence algorithm is proposed to search for novel architectures without human intervention that can achieve comparable performance to those of human-designed architectures. This work is inspired by current neural architecture search approaches based on reinforcement learning and genetic algorithm. However, not much attention is paid towards swarm intelligence metaheuristics-based neural architecture search. A framework is proposed for automatically designing neural architectures based on a swarm intelligence metaheuristic: Crow Search Algorithm. First, Crow Search Algorithm is integrated with binary network representation. To make it compatible for Neural Architecture Search, the original distance metric is replaced with hamming distance-based similarity measure. Second, the tuning parameters of Crow Search Algorithm are reduced by replacing the static flight length parameter with our dynamic flight length distribution algorithm. Third, the target selection method (random selection) is replaced by tournament select method. The proposed framework is used to search for architectures on MNIST, CIFAR10, and CIFAR100 datasets and achieved 0.18%, 3.48%, and 15.64% test error, respectively. Furthermore, small-scale transfer experiments are conducted to search architectures for Tiny ImageNet and achieved 34.43% test error. Nonparametric statistical analysis is performed to validate the impact of each modification in improving the quality of search space exploration. The proposed framework has achieved comparable performance with the state-of-the-art approaches, with a comparatively simpler approach and minimum human intervention. The proposed framework can be used to develop completely automated systems for designing architectures for various data-based classification applications.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2020.3031599