For approval of new courses and deletions or
modifications to an existing course.
Course Approval Form
registrar.gmu.edu/facultystaff/curriculum
Action Requested:
Course Level:
x Create new course
Inactivate existing course
Modify existing course (check all that apply)
Title
Prereq/coreq
Other:
College/School:
Submitted by:
Subject Code:
Credits
Schedule Type
Undergraduate
x Graduate
Repeat Status
Restrictions
Grade Type
Department:
Ext:
3-1693
VSE
Rajesh Ganesan
OR
Number:
Effective Term:
670
(Do not list multiple codes or numbers. Each course proposal must
have a separate form.)
Title:
SEOR
Email:
x
Fall
Spring
Summer
rganesan@gmu.edu
Year
2013
Current
Banner (30 characters max including spaces)
New
Metaheuristics for Optimization
Credits:
x
(check one)
Grade Mode:
Fixed
Variable
x
(check one)
3
Repeat Status:
x
Not Repeatable (NR)
Repeatable within degree (RD)
Repeatable within term (RT)
(check one)
Regular (A, B, C, etc.)
Satisfactory/No Credit
Special (A, B C, etc. +IP)
Prerequisite(s):
OR 441/541 or permission of instructor
Schedule Type:
(check one)
LEC can include
LAB or RCT
x
Maximum credits
allowed:
Lecture (LEC)
Lab (LAB)
Recitation (RCT)
Internship (INT)
Corequisite(s):
3
Independent Study (IND)
Seminar (SEM)
Studio (STU)
Instructional Mode:
x 100% face-to-face
Hybrid: ≤ 50% electronically delivered
100% electronically delivered
Restrictions Enforced by System: Major, College, Degree, Program, etc. Include Code.
Are there equivalent course(s)?
x Yes
No
If yes, please list SYST 670
Catalog Copy for NEW Courses Only (Consult University Catalog for models)
Description (No more than 60 words, use verb phrases and present tense)
Course on the theory and practice of metaheuristics, i.e. solution search techniques for solving
combinatorial optimization problems. It will introduce the theory, applications (scheduling in
manufacturing, transportation, and in other engineering and service industries), and computational
aspects of directly searching for solutions to solve computationally complex optimization problems
without a well-defined analytical model.
Indicate number of contact hours:
Hours of Lecture or Seminar per week: 3
When Offered: (check all that apply)
x Spring
Fall
Summer
Notes (List additional information for
the course)
Equivalent to SYST 670
Hours of Lab or Studio:
0
Approval Signatures
Ariela Sofer
12/5/12
Department Approval
Date
College/School Approval
Date
If this course includes subject matter currently dealt with by any other units, the originating department must circulate this proposal for review by
those units and obtain the necessary signatures prior to submission. Failure to do so will delay action on this proposal.
Unit Name
Unit Approval Name
Unit Approver’s Signature
Date
For Graduate Courses Only
Graduate Council Member
Provost Office
Graduate Council Approval Date
For Registrar Office’s Use Only: Banner_____________________________Catalog________________________________
revised 11/8/11
SCHOOL PROPOSAL TO THE GRADUATE COUNCIL
BY
VOLGENAU SCHOOL OF ENGINEERING
1. CATALOG DESCRIPTION
OR 670 Metaheuristics for Optimization
Credits: 3 (NR)
Course on the theory and practice of metaheuristics, i.e. solution search techniques for solving combinatorial optimization
problems. It will introduce the theory, applications (scheduling in manufacturing, transportation, and in other engineering and
service industries), and computational aspects of directly searching for solutions to solve computationally complex optimization
problems without a well-defined analytical model.
Equivalent to OR 670
Prerequisite(s): OR541 (or) permission of instructor.
Hours of Lecture or Seminar per week: 3
2.
JUSTIFICATION
(a) Course Objectives:
At the conclusion of this course the student will have learned the art of searching the solution space to solving
optimization problems that do not have a well-defined analytical model, and how and why metaheuristics may be
the only method that can solve many of the large scale optimization problems, which will otherwise need an
enormous computational time to obtain (near-) optimal solutions. The student will also have advanced knowledge of
solving optimization problems using heuristic algorithms that do not require mathematical models to describe them.
The course will use both Matlab and spread sheets to solve optimization problems using the metaheuristic search
techniques, however prior knowledge of Matlab or spread sheet use is not needed.
(b) Course Necessity:
The field of metaheuristics provides alternate approaches to solving many optimization and control problems. Large
scale optimization and control has become very important in recent years, which involve complexity and often do
not have well-defined models. In many cases, even with well-defined models, the issue often is the large
computational needs that make the well-known methods such as LP, IP, NLP, and DP methods inefficient. Under
such circumstances, one may want to look at the solutions space directly and search for solutions that yield the best
objective function value. This can be done using heuristic algorithms that search the solution space. Hence, it is
imperative that students are offered this subject.
(c) Relationship to Existing Courses:
This is a new course in the SEOR program that has been designed to provide a wealth of knowledge that is directly
applicable to the needs of applications that are complex, adaptable, and large scale. The course compliments the
fundamentals learnt in OR441/OR541 and introduces a new and powerful alternative to solve many optimization
problems that do not have well-defined models and has high computational needs.
3.
APPROVAL HISTORY
SEOR Department
Date: 12/7/12
VSE Graduate Committee Date:
Date: 12/7/12
4.
SCHEDULING
Every spring semester, starting spring 2013 and every spring thereafter.
Proposed Instructors: Rajesh Ganesan, Karla Hoffman, Bjorn Berg
5.
COURSE OUTLINE
(a) Syllabus
Week 1 Introduction
Week 2 S-metaheuristics: Metaheuristics Implementation: Tabu search, Job shop scheduling
Week 3 Simulated Annealing, ILS
Excel examples
Week 4 m machine n job, flow shop, 1 m n job with setup times, Knapsack, TSP, Dispatching rules
Week 5 Shifting bottleneck heuristic - completion time, job-shop scheduling, hospital scheduling, training matrix
Week 6 Shifting bottleneck heuristic -weighted tardiness, Flow shop –weighted, tardiness, set-covering problems, GRASP for
capacitated minimum spanning tree,
Guided Local search (GLS)
Week 7 Midterm
Week 8 P-metaheuristics introduction, Genetic Algorithm, Genetic programming, excel for GA
Week 9 Estimation distribution algorithms, scatter search, Flexible assembly scheduling: paced and unpaced lines with and without
buffer, minimum part set MPS, profile fitting and FFLL Flexible Flow Line Loading heuristic
Week 10 Economic lot sizing and scheduling, rotation schedules, Matlab examples
Week 11 FFS frequency fixing and sequencing heuristic, Ant colony optimization, PERT CPM
Week 12 Particle swarm Optimization, Reservation systems
Week 13 PSO, Bee colony optimization, Course and exam timetabling, train scheduling
Week 14 Workforce scheduling, artificial immune system, Airline Crew scheduling
Week 15 Hybrid, parallel metaheuristics
Week 16 Final exam (exam week)
(b) Required Reading and Reference Material
Text: Operations scheduling – Michael Pinedo and Xiuli Chao
Metaheuristics – El-Gazhali Talbi
(c) Student Evaluation Criteria
Mid-term:
30%
Homework:
15%
Project:
25%
Final Exam:
30%