Handwritten Digit Recognition Using Logistic Regression, SVM, KNN and CNN Algorithms

K. Sujana  Kumari; G.  Murali

Abstract

The style of handwriting varies from person to person. Handwritten numbers are not always the same size, orientation and width. To develop a system to understand this, the machine recognizes handwritten digit images and classifies them into 10 digits (from 0 to 9). The recognition of handwritten digits is a technology which is used for the automatic recognizing and detecting handwritten digital data through various deep lerning and machine learningmodels. This paper uses a different machine learning algorithms to improve productivity anda variety of models to reduce complexity. Machine Learning is a subset of Artificial Intelligence, actually Machine Learningapplications which learns from previous experiences and it automatically improves with the previous experiences. This paper is about recognizing handwritten digits from 0 to 9 from the familiar Modified National Institute of Standards and Technology (MNIST) dataset, then comparision takes place between machine learning algorithms like Support Vector Machine(SVM), Logistic Regression, K-NearestNeighbor (KNN) and deep learning algorithm like CNN.

Abstract

Keywords