Handwritten Digit Recognizer

Abstract

The acknowledgment of Handwritten characters and digits has consistently been a truly challenging errand on account of the numerous varieties of transcribed characters with various composing styles. Handwriting digit recognition systems are designed to transform handwritten digits into machine-readable representations. Handwritten Numeral Recognition plays an important role in postal automation services mainly in countries like India wherein more than one languages and scripts .The major objectives of this work is to create effective and reliable methods for accurately detecting numerals in order to make banking procedures more convenient and accurate. This sort of clever framework is applied in different fields: really look at handling, handling of structures, programmed handling of manually written responses to an assessment, and so on. This last application is the subject of this work. We use various machine learning algorithms to get the best accuracy for our result,we have use three type of algorithms to get best accuracy for our data set these are SUPPORT VECTOR MACHINE,NEURAL NETWORK,AND CONVOLUTIONAL NEURAL NETWORK we got the best result in CONVOLUTIONAL NEURAL NETWORK which is 98% accuracy.