Use the attached python notebook as a reference–> scania_failures.ipynb1. Train a neural network to predict the class variable 2. Display the learning curve (Accuracy vs num of epochs) for 4 values of learning rate (0.0001, 0.001, 0.01 and 0.1) 3. Test your model on the test set and report the accuracy 4. Display the confusion matrix 5. Further optimize the model by changing the number of nodes and layers to obtain the best true positive rateGoal is to improve the Area Under the Curve (AUC) of the ROC. The baseline is 0.77. Try to change the number of nodes, number of layers, learning rate and number of epochs to optimize the AUC. Let us see who can get the best AUC.