recognize handwritten digit and give the best output. This project is to classify an individual handwritten word so that handwritten digit can be translated to a digital form. A set of training data will be used for this purpose. The objective is to design and implement a digit recognition in java that will detect handwritten digits drawn on screen similar to the training data. In this project, a simple convolutional neural network [CNN] is created to classify handwriting digits using MNIST dataset using TensorFlow.
algorithm that can recognize the handwritten digits which are scanned and sent as input by the user. The goal of this paper is to observe the variation of different algorithms that can classify the handwritten digits using different hidden layers, various number of epochs and to make a comparison based on the accuracy. This experiment is performed using the Modified National Institute of Standards and Technology (MNIST) dataset.