Does intlinprog use cuts during its Branch and Bound Algorithm?

I´m currently working with the Function intlinprog from the optimization toolbox and wonder, if it uses cuts during its Branch and Bound Algorithm? I read the descriptions of the function https://www.mathworks.com/help/optim/ug/intlinprog.html and https://www.mathworks.com/help/optim/ug/mixed-integer-linear-programming-algorithms.html and i am still not sure, if it uses cuts during its BnB. There is a mention of cuts and even an option to tell the function what cuts should be used, but does that only apply to cuts for preprocessing? If i had to guess, i would say that it doesn´t use cuts in BnB, hence it is called Branch and Bound and not Branch and Cut, but i´m just not entirely certain of it. I´m fairly new to MILP, so if my questions sounds somewhat silly in face of the clear mention of BnB in intlinprog, then that´s on me, but i don´t get why one wouldn´t use cuts in BnB? Another question would be, if that function can be made to use multiple CPU cores?

I hope someone can give me some clarity in that regard. Thanks in advance

Answer 1

[Bruno Luong]

The document

CutMaxIterations: Number of passes through all cut generation methods before entering the branch-and-bound phase,

It seems pretty clear to me

[Christian Putz]

Thanks for the quick answer.

Yes i saw that, too. It definitely was clear to me, that there are cuts before the BnB I just didn´t know, if there are still cuts in the BnB algorithm. There was no mention of it, so i assumed there are no, but then again i don´t understand why there woudn´t be. As i said i´m fairly new to MILP solving, so i might simply not grasp why there are none. It just seemed odd to me and that rose some suspicion, if i am getting something wrong.

[Alan Weiss]

You can see a fairly complete algorithm description of intlinprog here.

Alan Weiss

MATLAB mathematical toolbox documentation