H MOEA Software Availability

H.1 Introduction

This appendix describes some of the main public-domain MOEA software that is currently available. This description includes the following information:

Name: Name of the software system Description: Any relevant information about the system Environment: Software environment for which the software is intended Language: Programming language in which the software is written Availability: Software availability

Original sources of each of the systems analyzed are cited, but readers should be aware of the fact that many of these systems are also mirrored at the EMOO repository located at (see under Software): http://delta.cs.cinvestav.mx/~ccoello/EMOO/ with a mirror at: http://www.lania.mx/~ccoello/EMOO

Note that although commercial software for multi-objective optimization also exists (e.g., iSIGHT, which is briefly described in Chapter 5), such systems are not included here, since we limited the contents of this appendix to public- domain software. 678 H MOEA Software Availability

Table H.1: MOEA Software Name Description Environment Availability Includes imple- Any platform Antonio J. Nebro Algorithms in mentations in supporting Java ([email protected]), or download Java (JMetal) Java of: NSGA-II at: http://neo.lcc.uma.es/metal/ [358], SPEA2 index.html [1712], PAES [858], OMOPSO [1546], AbYSS [1137], MOCell [1136]. It also includes several test functions Multiple- Includes: Pareto /; Andrzej Jaszkiewicz Objective Simulated An- Standard C++ ([email protected]) nealing [305, 306], compiler or download at: in C++ Serafini’s Multiple http://www-idss.cs.put.poznan.pl/ (MOMHLib++) Objective Simu- ~jaszkiewicz/MOMHLib/ lated Annealing [1416], Ulungu et al.’s Multiple Ob- jective Simulated Annealing [1564], Multiple Objec- tive Genetic Local Search (MOGLS) [753], Ishibuchi & Murata’s Multiple Objective Ge- netic Local Search [728], NSGA [1457], NSGA-II [347], Multiple Objective Multi- ple Start Local Search [760]; Full source code in C++; Fully doc- umented; Open architecture for new contributions Micro-GA for Includes several Linux (i386 & Gregorio Toscano Pulido Multiobjective test functions; Full SPARC), Solaris ([email protected]) or download Optimization Source Code or SunOS and at: http://delta.cs.cinvestav.mx/ G++ (GNU ~ccoello/EMOO/EMOOsoftware.html C++ Compiler) The Pareto Full source code in Linux/Unix; Joshua D. Knowles Archived Evo- C, including tools GNU C Com- ([email protected]) lution Strategy for statistical com- piler or download at: (PAES) parison; Test func- http://dbkgroup.org/knowles/multi/ tions and results ParEGO: an Full source code in Linux/Unix; Joshua D. Knowles algorithm for C, including test GNU C Com- ([email protected]) multiobjective functions piler or download at: optimization http://dbkgroup.org/knowles/parego/ of expensive functions MOCK: Mul- Full source code in Linux/Unix; Julia Handl ([email protected]) tiobjective C++ with a Java GNU C++ or download at: clustering with interface; includes Compiler http://dbkgroup.org/handl/mock/ automatic de- generators of syn- termination of thetic data sets the number of clusters H.1 Introduction 679

Table H.1: continued Name Description Environment Availability Test problems Data files and C Linux/Unix; Joshua D. Knowles generator for source code GNU C Com- ([email protected]) the Multiobjec- piler or download at: tive Quadratic http://dbkgroup.org/knowles/mQAP/ Assignment Problem Multi-Objective Graphical inter- Windows 95 or Tan Kay Chen ([email protected]) Evolutionary face; Constraint- better and Mat- or download at: Algorithm Tool- handling facilities; Lab 5.3 or better http://vlab.ee.nus.edu.sg/ box Allows incorpora- ~kctan/moea.htm tion of preferences or goals. Described in [1508] Nondominated Full source code Linux/Unix; Kalyanmoy Deb ([email protected]) sorting ge- in C with three Standard C or download at: netic algorithm unconstrained compiler http://www.iitk.ac.in/kangal/ (NSGA) test functions; soft.htm Supports binary, integer, real, and enumerated types for the design variables Nondominated Full source code Linux/Unix; Kalyanmoy Deb ([email protected]) sorting genetic in C with con- Standard C or download at: algorithm II strained test func- compiler http://www.iitk.ac.in/kangal/ (NSGA-II) tions; Supports soft.htm (there are sev- real-encoding (and eral revisions genetic operators) available) for the design variables ǫ-MOEA (there Full source code Linux/Unix; Kalyanmoy Deb ([email protected]) are revisions in C with con- Standard C or download at: available) strained test func- compiler http://www.iitk.ac.in/kangal/ tions; Supports soft.htm real-encoding (and genetic operators) for the design variables A C++ Library Full source code Linux/Unix; Xianming Chen for MOEAs in C++ of VEGA Standard C ([email protected]) or download [1390], SPEA compiler at: http://delta.cs.cinvestav.mx/ [1719], NPGA ~ccoello/EMOO/EMOOsoftware.html [686], NSGA [1457] and the Pareto Tree Searching Ge- netic Algorithm (PTSGA) [221]; Code tailored to solve multiob- jective knapsack problems Multi-Objective, MOEA that uses MatLab Mex- Evan J. Hughes Probabilis- noisy nondomi- file; Requires ([email protected]) tic Selection nated ranking Unix/Linux or download at: Evolutionary and MatLab http://delta.cs.cinvestav.mx/ Algorithms 5.3 or better; ~ccoello/EMOO/EMOOsoftware.html (MOPSEA) Documented in [700] Strength Pareto Full source code in Linux/Unix; Eckart Zitzler Evolutionary C++ Standard C++ ([email protected]) Algorithm compiler or download at: (SPEA) ftp://ftp.tik.ee.ethz.ch/pub/ people/zitzler/spea.cc 680 H MOEA Software Availability

Table H.1: continued Name Description Environment Availability Kit for Evolu- A software pack- Java Download at: tionary Algo- age for develop- http://ls11-www.cs.uni-dortmund.de/ rithms (KEA) ment, analysis and people/schmitt/Daten/Kea/kea.jsp application of mul- tiobjective evolu- tionary algorithms Platform and Public-domain Linux and Win- Download at: Programming implementa- dows; Standard http://www.tik.ee.ethz.ch/pisa/ Language Inde- tions of NSGA-II C compiler pendent Inter- [358], SPEA2 face for Search [1713] and the Algorithms Indicator Based (PISA) Evolutionary Al- gorithm [1711]; Test problems and performance measures Graphical User Tools for the Plaftorm inde- Download at: Interface for analysis of results pendent; Binary http://guimoo.gforge.inria.fr/ Multi-objective produced by a executables Optimization multi-objective available for (Guimoo) evolutionary Windows algorithm PARAllel and Tools for the Linux/Unix; Download at: DIStributed design of both GNU C++ http://www2.lifl.fr/~cahon/ Evolving Ob- serial and parallel compiler paradisEO/index.html jects (Par- (or distributed) adisEO) multi-objective evolutionary algorithms