dl

Deep learning on image dataset Usage of neural network model on an image dataset using Keras: In this project, I used Keras for preprocessing, loading the image dataset from MNIST, building a Sequential model based on dense neural network using RELU and softmax activation function, training the model and predicting the accuracy of the model on the dataset. 3 hidden layers with 512 nodes in the first and the second layer and 10 nodes in the third layer were employed. This model gave 97.78 % accuracy. This project is attached as SeqNN.ipynb.