Integrated Convolutional Neural Network and Genetic Algorithm Based Technique for Emnist Handwritten Character Recognition
International Journal of Advanced Research in Science Communication and Technology · 29.05.2026 · 900
DOI: 10.48175/ijarsct-35887 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Advanced Research in Science Communication and Technology, online erschienen am 29.05.2026. Verfasst von Balram Yadav und 2 weiteren Autor:innen. Die Arbeit nennt 17 Literaturangaben.
Abstract (Originalsprache)
In the landscape of modern computational research, pattern recognition has emerged as a sprawling domain, offering immense opportunities for academic contribution. This field encompasses a wide variety of data modalities, including audio signals, video streams, and digital images of handwritten digits or characters. This research specifically focuses on the domain of handwritten character recognition and classification. Given the inherent variability in individual handwriting styles and the vast diversity of global languages, there is a critical need for models capable of high-precision identification. While a multitude of deep learning and traditional machine learning frameworks have been developed to address this challenge with varying degrees of success, Convolutional Neural Networks (CNNs) have recently dominated the computer vision sector due to their efficacy in real-world applications. This study introduces an integrated methodology designed to overcome the limitations of individual models and improve overall classification performance. Specifically, we propose a hybrid architecture—CNN-GA—which synergizes the global search capabilities of Genetic Algorithms (GA) with the hierarchical feature processing of Convolutional Neural Networks (CNN). In this proposed framework, the GA is employed as an optimization tool to select the most relevant features from the data, which are then processed by the CNN for final classification. To validate this approach, we utilized the extensive EMNIST dataset, a benchmark collection comprising 145,600 handwritten samples of characters A through Z. By implementing the CNN-GA model, we achieved a superior recognition accuracy of 97.98%. This performance indicates a significant advancement over previous methodologies, demonstrating that the fusion of evolutionary optimization and deep learning provides a more robust solution for complex pattern recognition tasks. The results confirm that our proposed model sets a new benchmark for accuracy in the automated interpretation of handwritten data.
Zitieren
APA
Yadav, B., Gupta, J. B., & Shukla, A. (2026). Integrated Convolutional Neural Network and Genetic Algorithm Based Technique for Emnist Handwritten Character Recognition. International Journal of Advanced Research in Science Communication and Technology, 900. https://doi.org/10.48175/ijarsct-35887
Vancouver
Yadav B, Gupta JB, Shukla A. Integrated Convolutional Neural Network and Genetic Algorithm Based Technique for Emnist Handwritten Character Recognition. International Journal of Advanced Research in Science Communication and Technology. 2026:900. doi:10.48175/ijarsct-35887
BibTeX
@article{Yadav2026,
author = {Yadav, Balram and Gupta, Jyoti Bala and Shukla, Abhinav},
title = {Integrated Convolutional Neural Network and Genetic Algorithm Based Technique for Emnist Handwritten Character Recognition},
journal = {International Journal of Advanced Research in Science Communication and Technology},
year = {2026},
pages = {900},
doi = {10.48175/ijarsct-35887}
}
EUVIMED