MNIST ("Modified National Institute of Standards and Technology") is the de facto “hello world” dataset of computer vision. Since its release in 1999, this classic dataset of handwritten images has served as the basis for benchmarking classification algorithms.
A model to detect hand-written digits from images. This is a project based on the Kaggle Competetion "Digit Recognizer". I have built a strong image classification model using a Sequential model from Keras to classify hand-written digits. The images that I have used to train and test my model can be found at the following link. https://www.kaggle.com/c/digit-recognizer/data
My submission on Kaggle gave a score of 0.97928.The competition was evaluated on the categorization accuracy of the predictions submitted (the percentage of images you get correct).