Skip to content
Log in

Accuracy–Efficiency Analysis of Optimizer Variants and Convolutional Parameter Configurations for Handwritten Digit Recognition

Muhammad Munsarif Richmond Ampah-Mensah
Jul 2026 · Cureus Journal of Computer Science · Vol 3 · 0 citations · 15 references

TL;DR

The results show that peak accuracy and practical efficiency lead to different conclusions, and suggest that lightweight CNN design for handwritten digit recognition should be evaluated through accuracy-time trade-offs, not only through maximum accuracy.

Abstract

Handwritten digit recognition remains useful for studying compact image-classification models, although the MNIST (Modified National Institute of Standards and Technology) benchmark is no longer sufficient for claiming state-of-the-art novelty. Accordingly, this study frames MNIST as a controlled setting for analyzing accuracy-efficiency behavior in lightweight convolutional neural networks (CNNs). It examines how optimizer choice, epoch count, data augmentation, and convolutional parameter settings jointly affect classification accuracy and execution time. The experimental design covers two 5-layer CNN configurations, five parameter-varying cases, eight named optimizer settings, one baseline control setting, five epoch values, and two augmentation conditions. In total, 540 paired observations of accuracy and execution time are analyzed. The results show that peak accuracy and practical efficiency lead to different conclusions. The best observed accuracy was 99.50% for the standard CNN-5LS configuration with Nadam at 120 epochs after augmentation, but this required 1,011 s. RMSprop reached 99.47% after augmentation at both 15 and 120 epochs in the standard configuration, while the 15-epoch setting required only 114 s. In the parameter-varied CNN-5L model, Case-1 was generally the most efficient and stable configuration, while wider dilation and stride combinations frequently reduced reliability. Data augmentation improved weak or unstable settings more strongly than already stable configurations, and Ftrl did not produce meaningful learning in these experiments. These findings suggest that lightweight CNN design for handwritten digit recognition should be evaluated through accuracy-time trade-offs, not only through maximum accuracy. The scope is practical configuration guidance within the MNIST experimental setting.

DOI Semantic Scholar
View source

Similar papers

2026

Optimizing MNIST Digit Recognition for Edge Devices: A Multi-dimensional Evaluation Approach

The authors introduce the Edge Suitability Score (ESS), a composite metric that combines normalized accuracy, model size, and inference speed into a single value, weighted at 0.40, 0.35, and 0.25 to reflect their relative importance for microcontroller deployment.

Rajneesh Yadav, Aakash Parmar · 0 citations
Open access Jul 2026

Integrating Fuzzy Logic and Fractional-Order Operators in Convolutional Neural Networks for Handwritten Digits and Characters Recognition

This study focuses on enhancing handwritten Devanagari character recognition using deep learning models, specifically fine-tuned Convolutional Neural Networks (CNNs), combined with hybrid mathematical methods for image enhancement, proposing a fuzzy-enabled Power-Law transformation for image enhancement.

Akshara Sreenivasan, Vinodkumar Vinodkumar Arumugam, Sriramakrishnan Pathmanaban et al. · 0 citations
Review Open access Jul 2026

A Comparative Study of Deep Learning Models for Image Classification: Simple MLP, Deep MLP, Basic CNN, LeNet CNN

This study presents a comparative analysis of four deep learning architectures: Simple Multi-Layer Perceptron (MLP), Deep MLP, Basic Convolutional Neural Network (CNN), and LeNet Convolutional Neural Network (CNN). The models are evaluated on the MNIST handwritten digit dataset under identical experimental conditions using CPU-only hardware. Each model is assessed based on classification accuracy, training time, and number of trainable parameters. Experimental results demonstrate that convolutional architectures consistently outperform fully connected architectures in classification accuracy, with the Basic CNN achieving the highest validation accuracy of 99.15% and the LeNet CNN offering the best balance between performance and efficiency with only 61,706 trainable parameters. The findings confirm that architectural design has a greater influence on model performance than the number of layers or parameters alone. This study provides a comprehensive review of the relevant literature, detailed architectural descriptions, experimental methodology, and a thorough analysis of results to support the comparative conclusions.

Baffoe Nicholas, Jephthah Kwame Lanor · 0 citations
Open access 2020

A Comparative Study of Deep Learning Architectures for Image Classification

Image classification is a major area in computer vision, driven by rapid advances in deep learning. Over the last decade, convolutional neural networks (CNNs) and their variants have achieved high performance in applications such as medical diagnosis, autonomous driving, industrial inspection, remote sensing, and biometrics. However, choosing the right model remains challenging due to trade-offs between accuracy, computational cost, efficiency, and robustness. This paper presents a comparative study of different deep learning architectures, including classical CNNs, deep hierarchical models, residual and dense networks, and compound-scaled architectures. Using a common evaluation framework and standard datasets, the study analyzes performance based on key design factors such as depth, width, receptive field, skip connections, and normalization. Theoretical concepts like convolution operations, residual learning, and optimization are also discussed.The results show that deeper networks provide better representation, while residual connections and compound scaling improve training stability and efficiency. Lightweight models perform well in resource-limited and real-time environments. Overall, the study offers practical guidance for selecting suitable architectures and highlights future research areas such as neural architecture search, self-supervised learning, and efficient model deployment.

Riyaz Mohammed · 0 citations