A Basic Neural Network Trained on the infamous MNIST dataset.
This Neural Network takes input as 784 pixels, has a hidden layer of 10 Neurons and an Output layer of 10 Neurons representing each Digit. Having a Total of 7960 parameters.
Neural Network after being Trained have it's parameters stored into Dataset.json file to be able to extract the parameters and test the performance. During Testing it's performance result varies generally from 80% to 90% accuracy.