Dear students,

A new discussion topic has been opened on Moodle in the Neural Networks section.

Please answer the following question:

Why are activation functions essential in neural networks? What would happen if a neural network used only linear activation functions?

Your participation is important: it will help you better understand the real-world challenges encountered when training neural networks.

Please post at least one answer and, if possible, comment on a classmate's contribution.

We look forward to your contributions.