Python

$1.$ Revisiting transformation of data in preparation where scaling may be beneficial, explain the process. When should scaling be performed? Why do you not fit on the test dataset in supervised learning? Find a source and cite it that covers the topic and provide a summary of what can happen when you fit$-$transform the test set. Are there other transformations that should not be performed on the test set $($which$)?$

$2.$ Discuss the purpose of activation and optimization functions in neural networks.

$1.$ Discuss the pros and cons of Sigmoid and ReLu.

$2.$ What activation functions are used in the output layer and for what situations?

$3.$ Perform some documentation research on how the learning rate for a neural network may dynamically be adjusted with training operations. Cite the documentation and provide an explanation how to implement.