Should Machine Learning start with the classics of Pattern Recognition Theory?

Question

Machine learning

Hello.
There are books dated 70-80 years. For example (from what I found):

• Fukunaga -- Introduction to the statistical theory of pattern recognition
  
• Nilson -- Learning Machines
  
• Tu, Gonzalez -- Principles of Pattern Recognition
  

Objective: to master machine learning in order to properly understand and subsequently work in this area.
Q: How smart is it to start with these books? How relevant to my purposes are these books, or books published at the time?

Answer 1

What for? Can't you find newer books? Or is it just a matter of principle?
It is useful to read the classics, but based on the fact that something may become outdated during this time, and something new will appear, I would advise you to navigate the book no more than the last decade, and even less for neural networks.