This model uses the popular MNIST dataset to classify images. An Artificial Neural Network Model has been adopted with 2 dense layers and 'relu' which is Rectified Linear Unit is used as the activators. Training has been attained with an accuracy of 99.33%. The first 50 images with its predicted as well ground truth values have been shown in the end.
pavithra-rajan / image-classification Goto Github PK
View Code? Open in Web Editor NEWImage classification of handwritten numbers in MNIST dataset.