Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 56879 (19)
- University: Sharif University of Technology
- Department: Computer Engineering
- Advisor(s): Habibi, Jafar
- Abstract:
- The topic of operations research and optimization is widely used in various fields, from transportation industries to military industries. Researchers in these fields seek to improve answers to problems and methods. In computer science, optimization methods are widely used. One of the most important problems in computer science are NP-hard problems; These problems are inherently not solvable in polynomial time. Therefore, the use of innovative methods, despite not guaranteeing the optimality of the solution, provide a suitable solution to bring the result closer to the optimal solution. Integer linear programming problems are also considered as NP-hard problems. The branch-and-bound algorithm, which is based on tree search, is used in solving integer-combination linear programming problems. During the execution of this algorithm, several decisions are made to approach the optimal solution. Innovative methods are used in these decisions and can ultimately lead to obtaining the optimal solution or close to it. On the other hand, with the development of machine learning algorithms and their use in different fields, new research has been launched. In recent years, efforts have been made to use machine learning in combinatorial optimization. Various learning methods, including deep neural networks, graph neural networks, and reinforcement learning, have been applied to improving optimization problems. Although these learning methods are efficient, they require powerful GPUs and also take a long time to learn the problems. In the solvers of integer combination linear programming problems, various heuristic methods are used in the form of combination methods, which include several heuristics. In this research, a method will be presented that tries to integrate learning models inspired by hybrid methods. The aim of this method is to reduce the dimensions of the problem space and increase the efficiency of problem solving in less time
- Keywords:
- Machine Learning ; Combinatorial Optimization ; Branching Proccess ; Linear Programming ; Mixed Integer Linear Programming ; Branch and Bound Method
Digital Object List
-
محتواي کتاب
- view
Bookmark