A Study on Neural Network Pruning Techniques for Efficient Model Deployment
Global Journal of Engineering Innovations and Interdisciplinary Research · 30.12.2025 · 5 (6)
DOI: 10.33425/3066-1226.1179 ↗Kurz gefasst
Fachartikel in der Zeitschrift Global Journal of Engineering Innovations and Interdisciplinary Research, online erschienen am 30.12.2025. Verfasst von Prasadu Gurram und 3 weiteren Autor:innen.
Abstract (Originalsprache)
In this study, we explore the effectiveness of various neural network pruning techniques for optimizingmodel deployment in resource-constrained environments. The increasing demand for efficient neuralnetworks, particularly in edge devices and mobile applications, necessitates methods that reducemodel complexity without significantly compromising performance. We investigate five pruningstrategies—weight pruning, neuron pruning, structured pruning, unstructured pruning, and layerpruning—comparing their impact on model size, inference time, and accuracy. Our experimentalresults demonstrate that while all pruning methods achieve substantial reductions in model size andcomputational cost, they differ in their trade-offs between efficiency and accuracy. Structured andunstructured pruning emerge as particularly effective, offering a balance between model compactnessand performance retention, making them suitable for deployment on hardware-optimized architectures.These findings provide valuable insights into selecting appropriate pruning strategies for specificapplications, guiding the development of more efficient neural networks in practical scenarios.
Zitieren
APA
Gurram, P., Sindhura, Y., & Bandari, R. (2025). A Study on Neural Network Pruning Techniques for Efficient Model Deployment. Global Journal of Engineering Innovations and Interdisciplinary Research, 5(6). https://doi.org/10.33425/3066-1226.1179
Vancouver
Gurram P, Sindhura, Y., Bandari R. A Study on Neural Network Pruning Techniques for Efficient Model Deployment. Global Journal of Engineering Innovations and Interdisciplinary Research. 2025;5(6). doi:10.33425/3066-1226.1179
BibTeX
@article{Gurram2025,
author = {Gurram, Prasadu and Sindhura and Y. and Bandari, Ravi},
title = {A Study on Neural Network Pruning Techniques for Efficient Model Deployment},
journal = {Global Journal of Engineering Innovations and Interdisciplinary Research},
year = {2025},
volume = {5},
number = {6},
doi = {10.33425/3066-1226.1179}
}
EUVIMED