Conferences Papers

Conferences Papers (39)

Randomized Parameter Settings for Heterogeneous Workers in a Pool-Based Evolutionary Algorithm

by
Abstract

Recently, several Pool-based Evolutionary Algorithms (PEAs) have been proposed, that asynchronously distribute an evolutionary search among heterogeneous devices, using controlled nodes and nodes outside the local network, through web browsers or cloud services. In PEAs, the population is stored in a shared pool, while distributed processes called workers execute the actual evolutionary search. This approach allows researchers to use low cost computational power that might not be available otherwise. On the other hand, it introduces the challenge of leveraging the computing power of heterogeneous and unreliable resources. The heterogeneity of the system suggests that using a heterogeneous parametrization might be a better option, so the goal of this work is to test such a scheme. In particular, this paper evaluates the strategy proposed by Gong and Fukunaga for the Island-Model, which assigns random control parameter values to each worker. Experiments were conducted to assess the viability of this strategy on pool-based EAs using benchmark problems and the EvoSpace framework. The results suggest that the approach can yield results which are competitive with other parametrization approaches, without requiring any form of experimental tuning.

Published in
Parallel Problem Solving from Nature – PPSN XIII
Date of conference
September 13 - 17, 2014
Read more...

NEAT, There's No Bloat

by
Abstract

The Operator Equalization (OE) family of bloat control methods have achieved promising results in many domains. In particular, the Flat-OE method, that promotes a flat distribution of program sizes, is one of the simplest OE methods and achieves some of the best results. However, Flat-OE, like all OE variants, can be computationally expensive. This work proposes a simplified strategy for bloat control based on Flat-OE. In particular, bloat is studied in the NeuroEvolution of Augmenting Topologies (NEAT) algorithm. NEAT includes a very simple diversity preservation technique based on speciation and fitness sharing, and it is hypothesized that with some minor tuning, speciation in NEAT can promote a flat distribution of program size. Results indicate that this is the case in two benchmark problems, in accordance with results for Flat-OE. In conclusion, NEAT provides a worthwhile strategy that could be extrapolated to other GP systems, for effective and simple bloat control.

Published in
Lecture Notes in Computer Science. 17th European Conference, EuroGP 2014. 
Date of conference
April 23 - 25, 2014
Read more...

A Comparison of Fitness-Case Sampling Methods for Symbolic Regression with Genetic Programming

by
Abstract

The canonical approach towards fitness evaluation in Genetic Programming (GP) is to use a static training set to determine fitness, based on a cost function averaged over all fitness-cases. However, motivated by different goals, researchers have recently proposed several techniques that focus selective pressure on a subset of fitness-cases at each generation. These approaches can be described as fitness-case sampling techniques, where the training set is sampled, in some way, to determine fitness. This paper shows a comprehensive evaluation of some of the most recent sampling methods, using benchmark and real-world problems for symbolic regression. The algorithms considered here are Interleaved Sampling, Random Interleaved Sampling, Lexicase Selection and a new sampling technique is proposed called Keep-Worst Interleaved Sampling (KW-IS). The algorithms are extensively evaluated based on test performance, overfitting and bloat. Results suggest that sampling techniques can improve performance compared with standard GP. While on synthetic benchmarks the difference is slight or none at all, on real-world problems the differences are substantial. Some of the best results were achieved by Lexicase Selection and Keep Worse-Interleaved Sampling. Results also show that on real-world problems overfitting correlates strongly with bloating. Furthermore, the sampling techniques provide efficiency, since they reduce the number of fitness-case evaluations required over an entire run.

Published in
EVOLVE - A Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation V
Date of conference
July 1 - 4, 2014
Read more...

Evaluating the Effects of Local Search in Genetic Programming

by
Abstract

Genetic programming (GP) is an evolutionary computation paradigm for the automatic induction of syntactic expressions. In general, GP performs an evolutionary search within the space of possible program syntaxes, for the expression that best solves a given problem. The most common application domain for GP is symbolic regression, where the goal is to find the syntactic expression that best fits a given set of training data. However, canonical GP only employs a syntactic search, thus it is intrinsically unable to efficiently adjust the (implicit) parameters of a particular expression. This work studies a Lamarckian memetic GP, that incorporates a local search (LS) strategy to refine GP individuals expressed as syntax trees. In particular, a simple parametrization for GP trees is proposed, and different heuristic methods are tested to determine which individuals should be subject to a LS, tested over several benchmark and real-world problems. The experimental results provide necessary insights in this insufficiently studied aspect of GP, suggesting promising directions for future work aimed at developing new memetic GP systems.

Published in
EVOLVE - A Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation V
Date of conference
July 1 - 4, 2014
Read more...

Autonomous Demand-Side Management System Based on Monte Carlo Tree Search

by
Abstract

Smart Grid (SG) technologies are becoming increas- ingly dynamic, motivating the use of computational intelligence to support the SG by predicting and intelligently responding to certain requests (e.g., reducing electricity costs given fluctuating prices). The presented work intends to do precisely this, to make intelligent decisions to switch on electric devices at times when the electricity price (prices that change over time) is the lowest while at the same time attempting to balance energy usage by avoiding turning on multiple devices at the same time, whenever possible. To this end, we use Monte Carlo Tree Search (MCTS), a real-time decision algorithm. MCTS takes into consideration what might happen in the future by approximating what other entities/agents (electric devices) might do via Monte Carlo simulations. We propose two variants of this method: (a) maxn MCTS approach where the competition for resources (e.g., lowest electricity price) happens in one single decision tree and where all the devices are considered, and (b) two-agent MCTS approach, where the competition for resources is distributed among various decision trees. To validate our results, we used two scenarios, a rather simple one where there are no constraints associated to the problem, and another more complex, and realistic scenario with equality and inequality constraints associated to the problem. The results achieved by this real-time decision tree algorithm are very promising, specially those achieved by the maxn MCTS approach.

Published in
IEEE International Energy Conference (EnergyCon)
Date of conference
May 13 - 16, 2014
Read more...

Design of composite correlation filters for object recognition using multi-objective combinatorial optimization

by

  1. Alejandra Serrano Trujillo, Víctor H Díaz Ramírez and Leonardo Trujillo. Design of composite correlation filters for object recognition using multi-objective combinatorial optimization. 2013, 885605-885605-10. URL, DOI BibTeX

    @inproceedings{doi:10.1117/12.2024597,
    	author = "Serrano Trujillo, Alejandra and Díaz Ramírez, Víctor H. and Trujillo, Leonardo",
    	title = "Design of composite correlation filters for object recognition using multi-objective combinatorial optimization",
    	journal = "Proc. SPIE",
    	volume = 8856,
    	pages = "885605-885605-10",
    	year = 2013,
    	doi = "10.1117/12.2024597",
    	url = "http://dx.doi.org/10.1117/12.2024597"
    }
    

Abstract

Correlation filters for object recognition represent an attractive alternative to feature based methods. These filters are usually synthesized as a combination of several training templates. These templates are commonly chosen in an ad-hoc manner by the designer, therefore, there is no guarantee that the best set of templates is chosen. In this work, we propose a new approach for the design of composite correlation filters using a multi-objective evolutionary algorithm in conjunction with a variable length coding technique. Given a vast search space of feasible templates, the algorithm finds a subset that allows the construction of a filter with an optimized performance in terms of several performance metrics. The resultant filter is capable of recognizing geometrically distorted versions of a target in high cluttering and noisy conditions. Computer simulation results obtained with the proposed approach are presented and discussed in terms of several performance metrics. These results are also compared to those obtained with existing correlation filters.

Published in
Proc. SPIE 8856, Applications of Digital Image Processing XXXVI
Date of conference
26 September 2013
Read more...

Preliminary Study of Bloat in Genetic Programming with Behavior-Based Search

by
  1. Leonardo Trujillo, Enrique Naredo and Yuliana Martínez. Preliminary Study of Bloat in Genetic Programming with Behavior-Based Search. In Michael Emmerich, Andre Deutz, Oliver Schuetze, Thomas Bäck, Emilia Tantar, Alexandru-Adrian Tantar, Pierre Del Moral, Pierrick Legrand, Pascal Bouvry and Carlos A Coello (eds.). EVOLVE - A Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation IV. Advances in Intelligent Systems and Computing series, volume 227, Springer International Publishing, 2013, pages 293-305. URL, DOI BibTeX

    @incollection{,
    	year = 2013,
    	isbn = "978-3-319-01127-1",
    	booktitle = "EVOLVE - A Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation IV",
    	volume = 227,
    	series = "Advances in Intelligent Systems and Computing",
    	editor = "Emmerich, Michael and Deutz, Andre and Schuetze, Oliver and Bäck, Thomas and Tantar, Emilia and Tantar, Alexandru-Adrian and Moral, Pierre Del and Legrand, Pierrick and Bouvry, Pascal and Coello, Carlos A.",
    	doi = "10.1007/978-3-319-01128-8_19",
    	title = "Preliminary Study of Bloat in Genetic Programming with Behavior-Based Search",
    	url = "http://dx.doi.org/10.1007/978-3-319-01128-8_19",
    	publisher = "Springer International Publishing",
    	keywords = "Bloat; Genetic Programming; Novelty Search",
    	author = "Trujillo, Leonardo and Naredo, Enrique and Martínez, Yuliana",
    	pages = "293-305"
    }
    
Abstract

Bloat is one of the most interesting theoretical problems in genetic programming (GP), and one of the most important pragmatic limitations in the development of real-world GP solutions. Over the years, many theories regarding the causes of bloat have been proposed and a variety of bloat control methods have been developed. It seems that one of the underlying causes of bloat is the search for fitness; as the fitness-causes-bloat theory states, selective bias towards fitness seems to unavoidably lead the search towards programs with a large size. Intuitively, however, abandoning fitness does not appear to be an option. This paper, studies a GP system that does not require an explicit fitness function, instead it relies on behavior-based search, where programs are described by the behavior they exhibit and selective pressure is biased towards unique behaviors using the novelty search algorithm. Initial results are encouraging, the average program size of the evolving population does not increase with novelty search; i.e., bloat is avoided by focusing on novelty instead of quality.

Published in
EVOLVE - A Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation IV
Volume 227
Pages 293-305
http://link.springer.com/chapter/10.1007%2F978-3-319-01128-8_19
Date of conference
10-13  July 2013
ISSN
2194-5357
ISBN
978-3-319-01128-8
Read more...

Searching for novel clustering programs

by
  1. Enrique Naredo and Leonardo Trujillo. Searching for novel clustering programs. In GECCO. 2013, 1093-1100. BibTeX

    @inproceedings{DBLP:conf/gecco/NaredoT13,
    	author = "Enrique Naredo and Leonardo Trujillo",
    	title = "Searching for novel clustering programs",
    	booktitle = "GECCO",
    	year = 2013,
    	pages = "1093-1100",
    	ee = "http://doi.acm.org/10.1145/2463372.2463505",
    	crossref = "DBLP:conf/gecco/2013",
    	bibsource = "DBLP, http://dblp.uni-trier.de"
    }
    
  2. Christian Blum and Enrique Alba (eds.). Genetic and Evolutionary Computation Conference, GECCO '13, Amsterdam, The Netherlands, July 6-10, 2013. ACM, 2013. BibTeX

    @proceedings{DBLP:conf/gecco/2013,
    	editor = "Christian Blum and Enrique Alba",
    	title = "Genetic and Evolutionary Computation Conference, GECCO '13, Amsterdam, The Netherlands, July 6-10, 2013",
    	booktitle = "GECCO",
    	publisher = "ACM",
    	year = 2013,
    	isbn = "978-1-4503-1963-8",
    	ee = "http://dl.acm.org/citation.cfm?id=2463372",
    	bibsource = "DBLP, http://dblp.uni-trier.de"
    }
    
Abstract

Novelty search (NS) is an open-ended evolutionary algorithm that eliminates the need for an explicit objective function. Instead, NS focuses selective pressure on the search for novel solutions. NS has produced intriguing results in specialized domains, but has not been applied in most machine learning areas. The key component of NS is that each individual is described by the behavior it exhibits, and this description is used to determine how novel each individual is with respect to what the search has produced thus far. However, describing individuals in behavioral space is not trivial, and care must be taken to properly define a descriptor for a particular domain. This paper applies NS to a mainstream pattern analysis area: data clustering. To do so, a descriptor of clustering performance is proposed and tested on several problems, and compared with two control methods, Fuzzy C-means and K-means. Results show that NS can effectively be applied to data clustering in some circumstances. NS performance is quite poor on simple or easy problems, achieving basically random performance. Conversely, as the problems get harder NS performs better, and outperforming the control methods. It seems that the search space exploration induced by NS is fully exploited only when generating good solutions is more challenging.

Published in
GECCO '13 Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference
Pages 1093-1100
http://dl.acm.org/citation.cfm?id=2463505
Date of conference
03-05 Abril 2013
ISBN
978-1-4503-1963-8
Read more...

A behavior-based analysis of modal problems

by
  1. Leonardo Trujillo, Lee Spector, Enrique Naredo and Yuliana Mart\'ınez. A behavior-based analysis of modal problems. In GECCO (Companion). 2013, 1047-1054. BibTeX

    @inproceedings{DBLP:conf/gecco/TrujilloSNM13,
    	author = "Leonardo Trujillo and Lee Spector and Enrique Naredo and Yuliana Mart\'{\i}nez",
    	title = "A behavior-based analysis of modal problems",
    	booktitle = "GECCO (Companion)",
    	year = 2013,
    	pages = "1047-1054",
    	ee = "http://doi.acm.org/10.1145/2464576.2482682",
    	crossref = "DBLP:conf/gecco/2013c",
    	bibsource = "DBLP, http://dblp.uni-trier.de"
    }
    
  2. Christian Blum and Enrique Alba (eds.). Genetic and Evolutionary Computation Conference, GECCO '13, Amsterdam, The Netherlands, July 6-10, 2013, Companion Material Proceedings. ACM, 2013. BibTeX

    @proceedings{DBLP:conf/gecco/2013c,
    	editor = "Christian Blum and Enrique Alba",
    	title = "Genetic and Evolutionary Computation Conference, GECCO '13, Amsterdam, The Netherlands, July 6-10, 2013, Companion Material Proceedings",
    	booktitle = "GECCO (Companion)",
    	publisher = "ACM",
    	year = 2013,
    	isbn = "978-1-4503-1964-5",
    	ee = "http://dl.acm.org/citation.cfm?id=2464576",
    	bibsource = "DBLP, http://dblp.uni-trier.de"
    }
    
Abstract

Genetic programming (GP) has proven to be a powerful tool for (semi)automated problem solving in various domains. However, while the algorithmic aspects of GP have been a primary object of study, there is a need to enhance the understanding of the problems where GP is applied. One particular goal is to categorize problems in a meaningful way, in order to select the best tools that can possibly be used to solve them. This paper studies modal problems, a conceptual class of problems recently proposed by Spector at GECCO 2012. Modal problems are those for which a solution program requires different modes of operation for different contexts. The thesis of this paper is that modality, in this sense, is better understood by analyzing program performance in behavioral space. The behavior-based perspective is seen as part of a scale of different forms of analyzing performance; with a coarse view given by a global fitness value and a highly detailed view provided by the semantics approach. On the other hand, behavioral analysis is seen as a flexible approach where the context of a program's performance is considered at in a domain-specific manner. The experimental evidence presented here suggests that behavior-based search could allow a GP to find programs with disjoint behavioral structures, that can satisfy the requirements of each mode of operation of a modal problem.

Published in
GECCO '13 Companion Proceeding of the fifteenth annual conference companion on Genetic and evolutionary computation conference companion
Volume 7835
Pages 1047-1054
http://dl.acm.org/citation.cfm?id=2482682
Date of conference
03-05 Abril 2013
ISBN
978-1-4503-1964-5
Read more...

EvoSpace-i: a framework for interactive evolutionary algorithms

by
  1. Mario Garc\'ıa Valdez, Juan Julián Merelo Guervós, Leonardo Trujillo, Francisco Fernández Vega, José C Romero and Alejandra Mancilla. EvoSpace-i: a framework for interactive evolutionary algorithms. In GECCO (Companion). 2013, 1301-1308. BibTeX

    @inproceedings{DBLP:conf/gecco/ValdezMTVRM13,
    	author = "Mario Garc\'{\i}a Valdez and Juan Juli{\'a}n Merelo Guerv{\'o}s and Leonardo Trujillo and Francisco Fern{\'a}ndez de Vega and Jos{\'e} C. Romero and Alejandra Mancilla",
    	title = "EvoSpace-i: a framework for interactive evolutionary algorithms",
    	booktitle = "GECCO (Companion)",
    	year = 2013,
    	pages = "1301-1308",
    	ee = "http://doi.acm.org/10.1145/2464576.2482709",
    	crossref = "DBLP:conf/gecco/2013c",
    	bibsource = "DBLP, http://dblp.uni-trier.de"
    }
    
  2. Christian Blum and Enrique Alba (eds.). Genetic and Evolutionary Computation Conference, GECCO '13, Amsterdam, The Netherlands, July 6-10, 2013, Companion Material Proceedings. ACM, 2013. BibTeX

    @proceedings{DBLP:conf/gecco/2013c,
    	editor = "Christian Blum and Enrique Alba",
    	title = "Genetic and Evolutionary Computation Conference, GECCO '13, Amsterdam, The Netherlands, July 6-10, 2013, Companion Material Proceedings",
    	booktitle = "GECCO (Companion)",
    	publisher = "ACM",
    	year = 2013,
    	isbn = "978-1-4503-1964-5",
    	ee = "http://dl.acm.org/citation.cfm?id=2464576",
    	bibsource = "DBLP, http://dblp.uni-trier.de"
    }
    
Abstract

Currently, a large number of computing systems and user applications are focused on distributed and collaborative models for heterogeneous devices, exploiting cloud-based approaches and social networking. However, such systems have not been fully exploited by the evolutionary computation community. This work is an attempt to bridge this gap, and integrate interactive evolutionary computation with a distributed cloud-based approach that integrates with social networking for collaborative design of artistic artifacts. Such an approach to evolutionary art could fully leverage the concept of memes as an idea that spreads from person to person, within a computational system. In particular, this work presents EvoSpace-Interactive, an open source framework for the development of collaborative-interactive evolutionary algorithms, a computational tool that facilitates the development of interactive algorithms for artistic design. A proof of concept application is developed on EvoSpace-Interactive called Shapes that incorporates the popular social network Facebook for the collaborative evolution of artistic images generated using the Processing programming language. Initial results are encouraging, Shapes illustrates that it is possible to use EvoSpace-Interactive to effectively develop and deploy a collaborative system.

Published in
Proceedings of the fifteenth annual conference companion on Genetic and evolutionary computation conference companion (GECCO '13 Companion)
Volume 7834
Pages 121-132
http://link.springer.com/chapter/10.1007%2F978-3-642-36955-1_11
Date of conference
03-05 Abril 2013
ISSN
0302-9743
ISBN
978-3-642-36955-1
Read more...
Subscribe to this RSS feed