Great video and explaination, probably the best on youtube. What makes it great is the really thorough explanation of all steps. A lot of tutors too quickly assume that students are completely up to speed on all "preliminaries". Thank you!
Thank you sooo much Sergiy, I'm currently studying in germany and my Professor could'nt explain it half as understandable as you in 1 hour of time! You saved me a lot of nerves!
hello Professor, thank you so much for the great explanation! I have some questions, I would be grateful if you could take the time to answer them. 1. @28:00 Since you enumerated all the patterns, and used those patterns to see which pattern gives you the highest objective of CGSP, does not that defeat the whole purpose of not relying on storing the whole matrix A? 2. @35:13, should not we drop the exiting basic variable from the RMP? otherwise, at each iteration the RMP will get larger, which means A will keep growing?
1. Yes, you are right; in this toy example this is done for illustration purposes, since the complete LP is small enough to show it entirely. 2. You could do that, but that (keeping only basic variables in the RMP) would be one of the two extremes, the other one being keeping all the columns introduced in the process. In the first extreme case you may end up spending more time on solving the subproblem, whereas in the second one the RMP may become too large. In practice one usually tries to balance between the two extremes (say, drop a nonbasic column from the RMP if it does not reenter the basis for a certain number of iterations).
Hi thanks for the lesson! I have a question. What happens if the solution of the CGSP is not a pattern of the problem. Suppose you solve the CGSP with knapsack and the solution is not one of your column.
Can the number of columns being used actively increase? If the number of columns increases to 4, then matrix B would be a 4*3 matrix and it would be impossible to inverse?
@@sergiybutenko When I change the number of cuts (80, 50, 100) when I look at probabilities with a maximum of 3 waste, the number of pieces does not integer in the equation. Is it because my solution of the equation is wrong or because these probabilities are not enough for each number of pieces? do i have to look at more probabilities.
Will you consider to give online live training/courses. I would like to take because i have questions and need to understand also 2D cut and stock problems.
@@sergiybutenko Thanks for you reply. and woow it is long way :))) Hello from Turkiye :) I need to understand this type of question for my thesis and i am really struggling- suffering :) In youtube you are the only one , explain fluent and detailed but i have missing part in basic knowledge about integer linear programming. Now i am watching other videos of you to figure out dual problems. I think you used duality in this example. (at least you can give such small trick :D:D )
@@sergiybutenko Sir , i handled 1D cutting stock problem.Thanks for you help via video. At final calculation i used excel solver and got good results. Thanks again. But now i need to understand 2D cut and stock problems. Cutting of rectangular stock plate in identical rectangular small pieces. Do you have any suggestion about this problem. Where should i start ? Do you have any idea about such problems ? Thanks in advance.