This is a mini-project I completed under Coursera Project Network.
In the project, I have built a Neural Network model that classifies images and predicts the hand-written digits with an accuracy of 96%.
- Dataset Used - MNIST dataset of handwritten digits.
- Data preprocessing - Unrolled the array to vector and applied Data Normalisation (x-m)/s.
- Creating model - Used ReLU activation function for the input and the hidden layers and Softmax activation function for the output layer.
- Training the model - Used 3 epochs for training the model.
- Evaluation - Getting an accuracy of 96%.