As a student majored in software engineering, the school of mine did not offer many courses about this topic(except an introduction course). I have to learn it all by myself, and I usually focused on modern deep learning structures. Thus, I do not know the history, nor the development of traditional methods. And don't even mention what advantages/disadvantages they got.

Therefore, as my research interests gradually require stronger foundational theories, I decided to learn it all over and take some notes here.

I would be grateful if anyone read this could offer me some advice.

Here are the links to the notes:

Traditional Learning

• 0.Recommended Papers
• 1.Bayeian-view Methods
• 2.Frequentist-view Methods

Modern Deep Learning

• 3.Graph Neural Networks