Simple neural network built with numpy to see classification metrics.
- Choosing a small number of training inputs in each batch, and processes the error for each sample is accumulated.
- Invoke the back propagation following this.
- Classify the predictions using confusion matrix.
