Comparative Evaluation of Machine Learning Techniques for Handwritten Digit Classification Using the MNIST Dataset

Performance Comparison of Support Vector Machine and Decision Trees for the Classification of Handwritten Digits

Handwritten digit recognition remains one of the fundamental applications of machine learning and pattern recognition due to its widespread use in banking, postal services, document processing and intelligent information systems. Although deep learning approaches have demonstrated remarkable performance, classical machine learning algorithms such as Support Vector Machine (SVM) and Decision Tree (DT) remain attractive because of their simplicity, computational efficiency and interpretability. This study evaluates and compares the performance of Support Vector Machine and Decision Tree classifiers for handwritten digit recognition using the Modified National Institute of Standards and Technology (MNIST) dataset. The dataset comprises 70,000 grayscale images of handwritten digits (0–9), each with a resolution of 28 × 28 pixels. Image preprocessing involved normalization of pixel values and standard data preparation before model implementation. Both classifiers were implemented using Python and the Scikit-learn library under identical experimental conditions, and model performance was evaluated using accuracy, precision, recall and F1-score. Experimental results showed that the Support Vector Machine outperformed the Decision Tree across all evaluation metrics. The SVM achieved an accuracy of 94.32%, precision of 94.41%, recall of 94.32%, and an F1-score of 94.35%, whereas the Decision Tree recorded 87.36% accuracy, 87.42% precision, 87.36% recall and 87.35% F1-score. The findings indicate that SVM provides superior classification performance for handwritten digit recognition, while Decision Tree offers faster implementation and greater interpretability. The study concludes that SVM is more suitable for applications requiring high recognition accuracy, whereas Decision Tree remains appropriate for applications where computational simplicity and model transparency are prioritized

Handwritten Digit Recognition Using Deep Learning

The Handwritten Digit Recognition System is a machine learning and deep learning–based project developed to accurately identify handwritten numerical digits from input images. The main objective of this project is to recognize digits ranging from 0 to 9 by using image processing techniques and a Convolutional Neural Network (CNN) model trained on the MNIST dataset. In this system, the handwritten digit image is first captured and preprocessed through steps such as grayscale conversion, resizing, normalization, and noise reduction to improve prediction accuracy. The processed image is then passed to the trained CNN model, which extracts important features and classifies the digit into the corresponding numerical class. The project uses backpropagation for learning, Adam optimizer for efficient weight optimization, and Softmax activation function in the output layer for multi-class classification. The trained model provides high accuracy and fast prediction results, making the system suitable for real-time applications. This project demonstrates the practical implementation of deep learning in image recognition and can be further extended for applications such as automatic form processing, postal code recognition, and bank cheque digit identification.

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.

Handwritten Digit Recognition Based on Convolutional Neural Network and Its Application in Overlapping Image Processing

Automatic Classification of Clean and Noisy Offline Handwritten Signatures Using Hybrid Machine Learning

The automatic partitioning of handwritten signatures into clean and noisy is crucial to reliable authentication in documents. A new approach is presented here that integrates traditional handcrafted texture features such as Histogram of Oriented Gradients (HOG), Local Binary Patterns (LBP) and Gray-Level Co-occurrence Matrix (GLCM) with deep learning models. We have performed experiments on an offline signature dataset DocSign V2 comprising 2,400 grayscale signature images (50 users, equal number of clean and noisy signatures) created for this work. Eight conventional classifiers are tested on individual and combinations of features. The performance of a CNN built for the task and also four transfer learning models - EfficientNetB0, ResNet50, MobileNetV2, and DenseNet121 - are compared. The experiments show the best result of 97.5% was achieved with the HOG-LBP+AdaBoost combination. This is better than the best of deep learning models, MobileNetV2, which reached 95%. This demonstrates that, particularly for small datasets, carefully designed handcrafted features can be used to effectively classify clean and noisy signatures automatically. Moreover, the proposed system can also be used as a post-processing module to enhance signatures for other analysis or verification.

Comprehensive Analysis of Face Recognition using Machine Learning and Deep Learning Models

Face recognition is a cornerstone in computer vision with its applications covering emotion analysis, healthcare and human computer interaction. This paper gives a complete explanation and classification of face recognition models by categorizing into Machine learning (ML), Deep Learning (DL) and hybrid techniques. Traditional methods rely on manual features and statistical evaluations that are relevant, whereas DL methods used convolutional and transformer-based techniques for classification and feature extraction. The integration of both techniques (hybrid) with its benefits performs better flexibility and more consistency for different conditions. This survey presents current innovations such as fuzzy similarity measures, self-distillation methods, segmentation improved Convolutional Neural Network (CNNs) and securely preserving quaternion networks. Applications including masked face detections, light variation enhancement and emotion-aware music recommendation is specified. The comparative analysis of cited works indicates that while DL models achieve superior recognition accuracy (typically >90% on benchmarks like FER-2013), they demand extensive data and computational resources, resulting in higher inference latency compared to hybrid models. Additionally, privacy concerns are an important challenge in multimodal biometric systems. This paper shows research gap in existing methods and demonstrates the demand for face recognition systems that are fast, protect user privacy and adjust with different environments for practical real world scenarios.