Using Neural Networks to Build an Efficient Classification Model for Classifying Images in the CIFAR-10 Dataset

Yu Huang

2024

Abstract

Image Classification has been a hot topic in recent years, with computer vision becoming essential for many real-life scenarios in fields like health and security. This paper proposes a Convolutional Neural Network (CNN) to classify images into separate classes from the Canadian Institute for Advanced Research dataset (CIFAR-10), with the objectives of achieving high accuracy and low loss. The model is built with repeating convolutional, pooling, and Normalization layers and is optimized with algorithms like dropout, gradient descent and early stopping further maximizing efficiency and accuracy of the model. Results show a high accuracy of 91.2% and a low loss of 0.401 with validation data, suggesting that this model is reliable and precise. Overall, this study builds an efficient classification model using a Convolutional Neural network and is used to be tested on the CIFAR-10 dataset, and the results show such architecture is viable in real-life scenarios.

Paper Citation

in Harvard Style

Huang Y. (2024). Using Neural Networks to Build an Efficient Classification Model for Classifying Images in the CIFAR-10 Dataset. In Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-754-2, SciTePress, pages 149-153. DOI: 10.5220/0013511400004619

in Bibtex Style

@conference{daml24,
author={Yu Huang},
title={Using Neural Networks to Build an Efficient Classification Model for Classifying Images in the CIFAR-10 Dataset},
booktitle={Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2024},
pages={149-153},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013511400004619},
isbn={978-989-758-754-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Using Neural Networks to Build an Efficient Classification Model for Classifying Images in the CIFAR-10 Dataset
SN - 978-989-758-754-2
AU - Huang Y.
PY - 2024
SP - 149
EP - 153
DO - 10.5220/0013511400004619
PB - SciTePress