Department of Computer Science

Slides:



Advertisements
Similar presentations
Topic Outline ? Black-Box Optimization Optimization Algorithm: only allowed to evaluate f (direct search) decision vector x objective vector f(x) objective.
Advertisements

Non-dominated Sorting Genetic Algorithm (NSGA-II)
Instructor Neelima Gupta Table of Contents Approximation Algorithms.
Angers, 10 June 2010 Multi-Objective Optimisation (II) Matthieu Basseur.
O(N 1.5 ) divide-and-conquer technique for Minimum Spanning Tree problem Step 1: Divide the graph into  N sub-graph by clustering. Step 2: Solve each.
Great Theoretical Ideas in Computer Science for Some.
Approximation Algorithms for TSP
1 The TSP : Approximation and Hardness of Approximation All exact science is dominated by the idea of approximation. -- Bertrand Russell ( )
NP-Complete Problems Polynomial time vs exponential time
Minimum Spanning Trees Kun-Mao Chao ( 趙坤茂 ) Department of Computer Science and Information Engineering National Taiwan University, Taiwan
The Evolution Trees From: Computational Biology by R. C. T. Lee S. J. Shyu Department of Computer Science Ming Chuan University.
Approximation Algorithms: Concepts Approximation algorithm: An algorithm that returns near-optimal solutions (i.e. is "provably good“) is called an approximation.
1 NP-Complete Problems Polynomial time vs exponential time –Polynomial O(n k ), where n is the input size (e.g., number of nodes in a graph, the length.
1 Combinatorial Dominance Analysis Keywords: Combinatorial Optimization (CO) Approximation Algorithms (AA) Approximation Ratio (a.r) Combinatorial Dominance.
9-1 Chapter 9 Approximation Algorithms. 9-2 Approximation algorithm Up to now, the best algorithm for solving an NP-complete problem requires exponential.
9-1 Chapter 9 Approximation Algorithms. 9-2 Approximation algorithm Up to now, the best algorithm for solving an NP-complete problem requires exponential.
Network Optimization Problems: Models and Algorithms
Approximation Algorithms Motivation and Definitions TSP Vertex Cover Scheduling.
Approximation Algorithms Ola Svensson. Course Information Goal: – Learn the techniques used by studying famous applications Graduate Course FDD
1 NP-Complete Problems (Fun part) Polynomial time vs exponential time –Polynomial O(n k ), where n is the input size (e.g., number of nodes in a graph,
Approximation Algorithms
Multiobjective Optimization Athens 2005 Department of Architecture and Technology Universidad Politécnica de Madrid Santiago González Tortosa Department.
Optimal Arrangement of Ceiling Cameras for Home Service Robots Using Genetic Algorithms Stefanos Nikolaidis*, ** and Tamio Arai** *R&D Division, Square.
Programming & Data Structures
Algorithms for Network Optimization Problems This handout: Minimum Spanning Tree Problem Approximation Algorithms Traveling Salesman Problem.
On comparison of different approaches to the stability radius calculation Olga Karelkina Department of Mathematics University of Turku MCDM 2011.
The Traveling Salesman Problem Approximation
1 Approximation Through Scaling Algorithms and Networks 2014/2015 Hans L. Bodlaender Johan M. M. van Rooij.
Princeton University COS 423 Theory of Algorithms Spring 2001 Kevin Wayne Approximation Algorithms These lecture slides are adapted from CLRS.
Although this may seem a paradox, all exact science is dominated by the idea of approximation. Bertrand Russell Approximation Algorithm.
1 Genetic Algorithms and Ant Colony Optimisation.
Minimum Spanning Trees CS 146 Prof. Sin-Min Lee Regina Wang.
WK15. Vertex Cover and Approximation Algorithm By Lin, Jr-Shiun Choi, Jae Sung.
Multiobjective Optimization for Locating Multiple Optimal Solutions of Nonlinear Equation Systems and Multimodal Optimization Problems Yong Wang School.
Multi-objective Evolutionary Algorithms (for NACST/Seq) summarized by Shin, Soo-Yong.
Evolving RBF Networks via GP for Estimating Fitness Values using Surrogate Models Ahmed Kattan Edgar Galvan.
Neural and Evolutionary Computing - Lecture 9 1 Evolutionary Multiobjective Optimization  Particularities of multiobjective optimization  Multiobjective.
Optimization/Decision Problems Optimization Problems – An optimization problem is one which asks, “What is the optimal solution to problem X?” – Examples:
Evolutionary Computing Chapter 12. / 26 Chapter 12: Multiobjective Evolutionary Algorithms Multiobjective optimisation problems (MOP) -Pareto optimality.
Steiner Tree Problem Given: A set S of points in the plane = terminals
1 ParadisEO-MOEO for a Bi-objective Flow-Shop Scheduling Problem May 2007 E.-G. Talbi and the ParadisEO team
Approximation Algorithms by bounding the OPT Instructor Neelima Gupta
Instructor Neelima Gupta Table of Contents Introduction to Approximation Algorithms Factor 2 approximation algorithm for TSP Factor.
Power Magnetic Devices: A Multi-Objective Design Approach
L-Dominance: An Approximate-Domination Mechanism
EMIS 8373: Integer Programming
School of Computing Clemson University Fall, 2012
Hamiltonian Cycle and TSP
Hamiltonian Cycle and TSP
Haim Kaplan and Uri Zwick
Approximation Algorithms
E190Q – Project Introduction Autonomous Robot Navigation
Computability and Complexity
Approximation Algorithms for TSP
Minimum Spanning Trees
Multi-Objective Optimization for Topology Control in Hybrid FSO/RF Networks Jaime Llorca December 8, 2004.
Heuristic Optimization Methods Pareto Multiobjective Optimization
Department of Computer Science University of York
A New multi-objective algorithm: Pareto archived dds
Multi-Objective Optimization
Approximation Algorithms
ISEN 601 Location Logistics
Chen-Yu Lee, Jia-Fong Yeh, and Tsung-Che Chiang
Minimum Spanning Trees
RM-MEDA: A Regularity Model-Based Multiobjective Estimation of Distribution Algorithm BISCuit EDA Seminar
MOEA Testing and Analysis
Minimum Spanning Trees
A Better Approximation for Minimum Total Routing Path Clustering Problem in 2-D Underwater Sensor Networks Wei Wang, Donghyun Kim, and Weili Wu, A Better.
Approximation Algorithms
Minimum Spanning Trees
Presentation transcript:

Department of Computer Science and Information Technology بهینه‌یابی مسائل چندهدفه با استفاده از الگوریتم‌های تقریبی Multi-Objective Optimization Using Approximation Algorithms منصور داودي منفرد Mansoor Davoodi mdmonfared@iasbs.ac.ir آذر ماه 1395

Multi-Objective Optimization (MOO) Cost Comfort

Multi-Objective Optimization (MOO) Solution space c f2 Objective space Pareto-optimal f1

Multi-Objective Optimization (MOO) f1 f2

General Formulation of MOO

Goals in MOO f2 f2 f1 f1 Pareto-Optimality Diversity

Weighted Sum method

Weighted Sum method f2 f1

Weighted Sum method f2 f1

Multi-Objective Evolutionary Algorithms Non-dominated Sorting Genetic Algorithm-II Strength Pareto EA Rudolph’s Elitist MOEA Distance-Based Pareto GA Thermodynamical GA Pareto-Archived Evolution Strategy …. Benchmark Problems and Comparison Metrics

Advantages and disadvantages of EAs General framework Reasonable resource consumption Parallel evolution Disadvantages No guarantee in finding optimal solutions Need to user parameter setting

Approximation Algorithms Guaranteed to run in polynomial time for all instances. Guaranteed to find solution within ratio 𝛼 of optimum. Challenge. Need to prove a solution's value is close to optimum, without even knowing what optimum value is!

Approximation Algorithms-MCC

Approximation Algorithms-TSP Input: Graph G, a complete weighted graph. Output: Minimum weight tour that visits all the nodes C B A D F E G

Approximation Algorithms-TSP C B A D F E G Double MST C B A D F E G MST C B A D F E G TSP Tour MST<OPTTSP TSP tour ≤2MST 𝛼=2

Schematic view of approximation factor Minimization objective f f(Opt) 𝑓 𝑥 ≤𝛼. 𝑓(𝑂𝑝𝑡) Maximization objective f 𝑓 𝑥 ≥𝛼. 𝑓(𝑂𝑝𝑡) f(Opt)

Approximation of MOO What is the meaning of “MOO approximation” ? What is the difficulty of “MOO approximation” ? f2 f2 f1 f1

Approximation of MOO 𝑓 1 𝑥′ = 𝑓 1 𝑥 𝛼 𝑓 2 𝑥′ = 𝑓 2 𝑥 𝛽 Def. Let 𝛱 be a bi-objective minimization problem with objectives 𝑓 1 and 𝑓 2 . Solution x is an 𝛼,𝛽 -approximation solution for 𝛱 if there is no solution y such that 𝑓 1 𝑥 >𝛼 𝑓 1 𝑦 and 𝑓 2 𝑥 ≥𝛽 𝑓 2 𝑦 , or 𝑓 1 𝑥 ≥𝛼 𝑓 1 𝑦 and 𝑓 2 𝑋 >𝛽 𝑓 2 𝑦 , f2 f1 𝑓 1 𝑥′ = 𝑓 1 𝑥 𝛼 𝑓 2 𝑥′ = 𝑓 2 𝑥 𝛽 x 𝛽 𝛼 x'

Approximation of MOO D𝐞𝐟. 𝜖-approximate Pareto curve is a set of solutions which are not dominated by any other by a ratio of more than 1+ 𝜖. f2 f1 𝜖-approximate Pareto curve P𝜖 is a set of solutions s such that there is no other solution 𝑠 ′ such that, for all s∈P𝜖, 𝑓 𝑖 𝑠 ′ ≥ 𝑓 𝑖 𝑠 1+𝜖 for some 𝑖 PType equation here.

Multi-objective optimization (MOO)

Multi-objective optimization (MOO) Optimal algorithm for Manhattan and 2 , 2 -approximation for the Euclidean metric 𝜃(n log n)

Bi-objective Path Planning

Bi-objective Path Planning

Bi-objective Path Planning

Bi-objective Path Planning

Bi-objective Path Planning 2 ,1 -approximation of the Euclidean path

Challenges … . Clustering Locating Planning Networking The real world is a trade-off world and it is full of conflict objectives. Any application of MOO approximation is welcome. Basic definitions of approximation Pareto-optimal curve. Complexity of MOO approximation. … . Clustering Locating Planning Networking

References I. Diakonikolas, Approximation of Multiobjective Optimization Problems, PhD thesis, Columbia University, 2011. M. Yannakakis, CH. Papadimitriou, On the approximability of offs - trade and optimal access of web source, 2000. V. Roostapour, I. Kiarazm, M. Davoodi, Deterministic Algorithm for 1-Median 1- Center Two-Objective Optimization Problem. TTCS, Springer, 2016. M. Davoodi, Bi-objective Path Planning using Deterministic Algorithms, submitted to Robotics and Autonomous Systems, March 2016. C. C. A. Coello, D. A. Van Veldhuizen, G. B. Lamont, Evolutionary algorithms for solving multi-objective problems. New York, 2002.

Thanks… ؟؟؟ Any Question ???