## PISA - A Platform and Programming Language Independent Interface for Search Algorithms (2003)

Citations: | 73 - 8 self |

### BibTeX

@INPROCEEDINGS{Bleuler03pisa-,

author = {Stefan Bleuler and Marco Laumanns and Lothar Thiele and Eckart Zitzler},

title = {PISA - A Platform and Programming Language Independent Interface for Search Algorithms},

booktitle = {},

year = {2003},

pages = {494--508},

publisher = {Springer}

}

### Years of Citing Articles

### OpenURL

### Abstract

This paper int roduces at ext based int rface (PISA)t hat allows t separat ty algorit hm-specific part of an op t mizer fromt he applicat ionspecific part . These part s are implement ed as independent programs forming freelycombinable modules.

### Citations

518 |
Evolutionary Algorithms for Solving Multi-Objective Problems. 2007
- Coello, Lamont, et al.
(Show Context)
Citation Context ...extension of this model was proposed in [11], which allows different sizes for parent and offspring population. However, this model does not consider an archive as an active part of the algorithm. In =-=[4]-=-, eight operational steps were proposed that an effective multiobjective evolutionary algorithm should incorporate as a generic structure. The authors remarked that the succession of these steps could... |

406 | L.: SPEA2: improving the strength pareto evolutionary algorithm for multiobjective optimization, Research Report
- Zitzler, Thiele
- 2001
(Show Context)
Citation Context ... the case of distributed processes, no communication error was ever detected. In a second experiment, a large application written in Java was tested with the well known multiobjective optimizer SPEA2 =-=[12]-=- written in C++ using the library TEA[5]. The purpose of the optimization was the design space exploration of a network processor including architecture selection, binding of tasks and scheduling, see... |

286 | Multiobjective evolutionary algorithms: analyzing the state-of-the-art." Evol Comput 8(2
- Veldhuizen, Lamont
- 2000
(Show Context)
Citation Context ...f (stochastic) operators that represent the ’genetic’ or ’evolutionary’ operations, e.g., in the universal evolutionary algorithm given in [2]. Amultiobjective extension of this model was prop=-=osed in [11]-=-, which allows different sizes for parent and offspring population. However, this model does not consider an archive as an active part of the algorithm. In [4], eight operational steps were proposed t... |

210 | Evolutionary computation: comments on the history and current state
- Back, Hammel, et al.
- 1997
(Show Context)
Citation Context ... formulation. The components of these models mainly consist of (stochastic) operators that represent the ’genetic’ or ’evolutionary’ operations, e.g., in the universal evolutionary algorithm g=-=iven in [2]-=-. Amultiobjective extension of this model was proposed in [11], which allows different sizes for parent and offspring population. However, this model does not consider an archive as an active part of ... |

44 | Running time analysis of multi-objective evolutionary algorithms on a simple discrete optimization problem
- Laumanns, Thiele, et al.
- 2002
(Show Context)
Citation Context ... In a first set of experiments, an interface has been written in the programming language C and extended with the simple multi-objective optimizer SEMO (selector) and the LOTZ problem (variator), see =-=[6]-=-. They have been tested on various platforms (Windows, LINUX, Solaris) where the two processes have been residing as well on different machines as on the same machine. Despite of the fact, that a shar... |

37 | Multiobjective evolutionary algorithms and applications
- Tan, Khor, et al.
- 2005
(Show Context)
Citation Context ...EO [1], but except for the TEA library none has a support for multiobjective optimization. Asecond group consists of implementations in MATLAB tool-boxes, for which also multiobjective versions exist =-=[9]-=-. These programming frameworks are characterized by a modular description of the algorithms, where different operators can be combined. Whenever a non-standard problem representation is used with sear... |

30 | L.: A unified model for multiobjective evolutionary algorithms with elitism
- Laumanns, Zitzler, et al.
- 2000
(Show Context)
Citation Context ...lties in finding a common structure for the succession of the different operations has led to a more coarsely grained formulation in the Unified Model of Multiobjective Evolutionary Algorithms (UMMEA =-=[7]-=-). This model was, however, not primarily designed from a software-technological viewpoint for easy implementation, but rather to facilitate comparative case studies on the effects of different operat... |

7 |
JEO: Java evolving objects
- Arenas, Dolin, et al.
- 2002
(Show Context)
Citation Context ...mentation effort of evolutionary algorithms. Most software engineering solutions resulted in (object oriented) class libraries, mainly written in C++, e.g. GAlib, EO [8] or TEA [5], or Java, e.g. JEO =-=[1]-=-, but except for the TEA library none has a support for multiobjective optimization. Asecond group consists of implementations in MATLAB tool-boxes, for which also multiobjective versions exist [9]. T... |

5 |
Using the CPLEX callable library
- Bixby, Boyed
- 1996
(Show Context)
Citation Context ...at can be used by engineers without any specific knowledge about the implementation of the opti3smization algorithms. Such approaches exist outside the evolutionary algorithms domain, e.g., the CPLEX =-=[3]-=- package for solving linear programming problems. For this purpose it is helpful to explicitly define the underlying model. Formal models for evolutionary algorithms do exist, as will be described in ... |

4 | TEA - a C++ library for the design of evolutionary algorithms
- Emmerich, Hosenberg
- 2000
(Show Context)
Citation Context ...mming process and implementation effort of evolutionary algorithms. Most software engineering solutions resulted in (object oriented) class libraries, mainly written in C++, e.g. GAlib, EO [8] or TEA =-=[5]-=-, or Java, e.g. JEO [1], but except for the TEA library none has a support for multiobjective optimization. Asecond group consists of implementations in MATLAB tool-boxes, for which also multiobjectiv... |

2 |
Easea comparisons on test functions : Galib versus eo
- Lutton, Collet, et al.
(Show Context)
Citation Context ... the programming process and implementation effort of evolutionary algorithms. Most software engineering solutions resulted in (object oriented) class libraries, mainly written in C++, e.g. GAlib, EO =-=[8]-=- or TEA [5], or Java, e.g. JEO [1], but except for the TEA library none has a support for multiobjective optimization. Asecond group consists of implementations in MATLAB tool-boxes, for which also mu... |

1 |
TEA: Toolbox for evolutionary algoritms – library documentation
- Biermann, Emmerich, et al.
- 1997
(Show Context)
Citation Context ...there are cases where standard representations such as binary strings and associated variation operators are adequate. For these situations, standard libraries are available to ease programming, e.g. =-=[1, 4]-=-. In summary, it is the task of the application engineer to define appropriate representations and neighborhood structures. In contrary, most optimizers in the multiobjective field work with a selecti... |