MIT OpenCourseWare


» 進階搜尋
 課程首頁
 教學大綱
 教學時程
 相關閱讀資料
 課堂講稿
 複習/實習課程
 作業
 測驗
 討論群組
 下載課程

教學大綱


本頁翻譯進度

燈號說明

審定:無
翻譯:蘇俊(簡介並寄信)
編輯:陳玉侖(簡介並寄信)


課程大綱

本課程對整數最佳化的理論、演算法和應用做了全面的介紹,共分為四部分:

第一部分:公式化和鬆弛,第1-8課

討論如何將整數最佳化問題公式化,如何增強公式以提高鬆弛的品質,如何獲得理想公式、整數最佳化的對偶性,以及如何在實際中和理論上解決因而發生的鬆弛。

第二部分:整數最佳化的幾何學和代數學,第9-14課

逐步闡明格論,概述影響整數最佳化的代數幾何學思想,並討論整數最佳化的幾何學。這些課堂講授為發展演算法提供基礎。

第三部分:整數最佳化的演算法,第15-22課

闡述割平面方法、積分的基本方法、枚舉和啟發式方法和逼近演算法。我們處理方法的主要特點是根據演算和幾何運算法二所自然發展出的演算方法。

第四部分:整數最佳化的延伸,第23-25課

處理混合整數最佳化和強健離散最佳化。這兩個領域在實際應用中都非常重要,因為現實世界通常都同時有連續和離散變數,並有著不確定的元素,這需要用易處理的方法來解決。

講師

Dimitris Bertsimas教授

指定教科書

Bertsimas, Dimitris和Robert Weismantel.《整數最佳化》,Belmont, Massachusetts: 動態觀點出版社,2004年12月. ISBN: 0975914626.這是這本書正式出版前的版本。

要求

項目 百分比
期末測驗 30%
期中測驗 30%
家庭作業 30%
課堂貢獻 (課堂參與及書評) 10%

Course Outline

The course is a comprehensive introduction to the theory, algorithms and applications of integer optimization and is organized in four parts:

Part I: Formulations and Relaxations, Lectures 1-8

Discusses how to formulate integer optimization problems, how to enhance the formulations to improve the quality of relaxations, how to obtain ideal formulations, the duality of integer optimization and how to solve the resulting relaxations both practically and theoretically.

Part II: Algebra and Geometry of Integer Optimization, Lectures 9-14

Develops the theory of lattices, outlines ideas from algebraic geometry that have had an impact on integer optimization, and discusses the geometry of integer optimization. These lectures provide the building blocks for developing algorithms.

Part III: Algorithms for Integer Optimization, Lectures 15-22

Develops cutting plane methods, integral basis methods, enumerative and heuristic methods and approximation algorithms. The key characteristic of our treatment is that our development of the algorithms is naturally based on the algebraic and geometric developments of Part II.

Part IV: Extensions of Integer Optimization, Lectures 23-25

Treats mixed integer optimization and robust discrete optimization. Both areas are practically significant as real world problems have very often both continuous and discrete variables and have elements of uncertainty that need to be addressed in a tractable manner.

Instructor

Professor Dimitris Bertsimas

Required Book

Bertsimas, Dimitris, and Robert Weismantel. Optimization over Integers. Belmont, Massachusetts: Dynamic Ideas, December 2004. ISBN: 0975914626. The book is the prepublication edition of the book.

Requirements

ACTIVITIES PERCENTAGES
Final Exam 30%
Midterm Exam 30%
Homework 30%
Contributions to Class (Class participation and comments on book) 10%

 
MIT Home
Massachusetts Institute of Technology Terms of Use Privacy