CP7027 MULTI OBJECTIVE OPTIMIZATION TECHNIQUES SYLLABUS FOR ME 3RD SEM CSE

ANNA UNIVERSITY, CHENNAI

REGULATIONS - 2013
CP7027 MULTI OBJECTIVE OPTIMIZATION TECHNIQUES SYLLABUS

ME 3RD SEM COMPUTER SCIENCE AND ENGINEERING SYLLABUS

CP7027 MULTI OBJECTIVE OPTIMIZATION TECHNIQUES SYLLABUS

OBJECTIVES:

 Learn fundamental principles of Multiobjective Optimization (MOP)

 Survey different Multiobjective Optimization algorithms

 Introduce various design issues of MOP

 Develop and Evaluate MOP Algorithms

 Learn Parallel and hybrid MOP Algorithms

 Learn other Metaheuristics

UNIT I INTRODUCTION AND CLASSICAL APPROACHES

Multiobjective optimization: Introduction - Multiobjective optimization problem-principles – Difference between single and multiobjective optimization – Dominance and Pareto Optimality , Classical Methods – Weighted Sum - Constraint method – Weighted Metric methods – Benson’s method - Value Function - Goal Programming methods – Interactive Methods

UNIT II MOP EVOLUTIONARY ALGORITHMS

Generic MOEA - Various MOEAs: MOGA, NSGA-II, NPGA, PAES, SPEA2, MOMGA, micro GA - Constrained MOEAs: Penalty Function approach - Constrained Tournament – Ray – Tai –Seow’s Method.

UNIT III THEORETICAL ISSUES

Fitness Landscapes - Fitness Functions - Pareto Ranking - Pareto Niching and Fitness Sharing - Recombination Operators - Mating Restriction - Solution Stability and Robustness - MOEA Complexity - MOEA Scalability - Running Time Analysis - MOEA Computational Cost - No Free Lunch Theorem.

UNIT IV MOEA TESTING, ANALYSIS, AND PARALLELIZATION

MOEA Experimental Measurements – MOEA Statistical Testing Approaches – MOEA Test Suites - MOEA Parallelization: Background – Paradigms – Issues - MOEA Local Search Techniques.

UNIT V APPLICATIONS AND ALTERNATIVE METAHEURISTICS

Scientific Applications: Computer Science and Computer Engineering - Alternative Metaheuristics: Simulated Annealing – Tabu Search and Scatter Search – Ant System – Distributed Reinforcement Learning – Particle Swarm Optimization – Differential Evolution – Artificial Immune Systems - Other Heuristics.

TOTAL:45 PERIODS

OUTCOMES:

Upon Completion of the course,students will be able to

 Explain MOP principles

 Explain classical methods to solve MOP problems

 Be familiar with and explain structures of different MOP algorithms

 Solve constrained MOP problems

 Explain various design issues of MOP algorithms

 Perform a evaluation and analysis of MOP algorithm results

 Explain parallelization of MOP algorithms

 Develop parallel and hybrid MOP algorithms

 Identify various real time MOP applications

 Explain other search algorithms

REFERENCES:

1. Carlos A. Coello Coello, Gary B. Lamont, David A. Van Veldhuizen, “Evolutionary

Algorithms for Solving Multi-objective Problems”, Second Edition, Springer, 2007.

2. Kalyanmoy Deb, “ Multi-Objective Optimization Using Evolutionary Algorithms”, John Wiley, 2002.

3. Aimin Zhoua, Bo-Yang Qub, Hui Li c, Shi-Zheng Zhaob, Ponnuthurai Nagaratnam Suganthan b, Qingfu Zhangd, “Multiobjective evolutionary algorithms: A survey of the state of

the art”, Swarm and Evolutionary Computation (2011) 32–49.

4. E Alba, M Tomassini, “Parallel and evolutionary algorithms”, Evolutionary Computation, IEEE Transactions on 6 (5), 443-462.

5. Crina Grosan, Ajith Abraham, “Hybrid Evolutionary Algorithms: Methodologies, Architectures, and Reviews”, Studies in Computational Intelligence, Vol. 75, Springer, 2007.

6. Christian Blum and Andrea Roli. 2003. Metaheuristics in combinatorial optimization: Overview and conceptual comparison. ACM Comput. Surv. 35, 3 (September 2003), 268-308.