This notebook is trained to classify the images of the digits that are handwritten using ML techniques.
MNIST database (stands for Modified National Institute of Standards and Technology) consists of large no. of handwritten digit images which can be used to train a model for classifying the digits.
LeNet - 5 is a CNN (Convolutional neural network) architecture which consist of 7 layers on whole (excluding input layer). This notebook is inspired from this architecture. The architecture that this notebook has is similar to LeNet - 5 but not same. The layers used in this notebooks are in this way, [CONV -> AVGPOOL -> CONV -> AVGPOOL -> DENSE -> DENSE -> SOFTMAX].
The test accuracy for this model is 98.057%
For the reference of the same code on kaggle, click on the button below...
