Handwritten Digit Recognition Accuracy Comparison Using Knn,Cnn And Svm.

Abstract

This project Handwritten digit recognition is a vital area in the field of computer vision. This paper compares the performance of convolutional neural networks (CNNs), k-nearest neighbors (KNN), and support vector machines (SVMs) for digit classification using the MNIST dataset. The dataset consists of 60,000 labeled images of handwritten digits 0-9 for training and 10,000 images for testing. Experimental results demonstrate that CNNs achieve the highest test accuracy of 98.6%, outperforming KNN (96.8%) and SVM (97.1%). The multi-layer feature extraction capability and depth of CNN models are better at capturing the unique visual patterns of different digit classes in the images. The study provides quantitative results that illustrate that CNN architectures yield state- of-the-art performance on MNIST digit classification compared to other common approaches