scispace - formally typeset
Search or ask a question

Showing papers in "IEEE Computational Intelligence Magazine in 2006"


Journal ArticleDOI
TL;DR: The introduction of ant colony optimization (ACO) is discussed and all ACO algorithms share the same idea and the ACO is formalized into a meta-heuristics for combinatorial problems.
Abstract: The introduction of ant colony optimization (ACO) and to survey its most notable applications are discussed. Ant colony optimization takes inspiration from the forging behavior of some ant species. These ants deposit Pheromone on the ground in order to mark some favorable path that should be followed by other members of the colony. The model proposed by Deneubourg and co-workers for explaining the foraging behavior of ants is the main source of inspiration for the development of ant colony optimization. In ACO a number of artificial ants build solutions to an optimization problem and exchange information on their quality through a communication scheme that is reminiscent of the one adopted by real ants. ACO algorithms is introduced and all ACO algorithms share the same idea and the ACO is formalized into a meta-heuristics for combinatorial problems. It is foreseeable that future research on ACO will focus more strongly on rich optimization problems that include stochasticity.

2,270 citations


Journal ArticleDOI
TL;DR: This article provides a general overview of the field now known as "evolutionary multi-objective optimization," which refers to the use of evolutionary algorithms to solve problems with two or more (often conflicting) objective functions.
Abstract: This article provides a general overview of the field now known as "evolutionary multi-objective optimization," which refers to the use of evolutionary algorithms to solve problems with two or more (often conflicting) objective functions. Using as a framework the history of this discipline, we discuss some of the most representative algorithms that have been developed so far, as well as some of their applications. Also, we discuss some of the methodological issues related to the use of multi-objective evolutionary algorithms, as well as some of the current and future research trends in the area.

1,213 citations


Journal ArticleDOI
TL;DR: Although still relatively young, the artificial immune system (AIS) is emerging as an active and attractive, field involving models, techniques and applications of greater diversity.
Abstract: During the last decade, the field of artificial immune system (A1S) is progressing slowly and steadily as a branch of computational intelligence (CI). There has been increasing interest in the development of computational models inspired by several immunological principles. In particular, some are building models mimicking the mechanisms in the biological immune system (BIS) to better understand its natural processes and simulate its dynamical behavior in the presence of antigens/pathogens. Most of the AIS models, however, emphasize designing artifacts - computational algorithms, techniques using simplified models of various immunological processes and functionalities. Like other biologically-inspired techniques, such as artificial neural networks, genetic algorithms, and cellular automata, AISs also try to extract ideas from the BIS in order to develop computational tools for solving science and engineering problems. Although still relatively young, the artificial immune system (AIS) is emerging as an active and attractive, field involving models, techniques and applications of greater diversity

499 citations


Journal ArticleDOI
TL;DR: Aocioscope is built to track the behavior of groups of humans in great detail and during long periods of time and finds that an amazing amount of human behavior can be accurately modeled by considering only certain sorts of mimicry, non-linguistic social signals, and stimulus-response mechanisms.
Abstract: Computational models of human intelligence almost always take the individual as the fundamental unit of analysis. But is this correct? For instance, if intelligence were primarily a group (or more generally, social network) phenomenon, then we would have to dramatically change how we approach the problem of computational intelligence. To explore this question, the author has built a `socioscope' to track the behavior of groups of humans in great detail and during long periods of time, What he finds is that an amazing amount of human behavior can be accurately modeled by considering only certain sorts of mimicry, non-linguistic social signals, and stimulus-response mechanisms. That is, human intelligence is substantially, and perhaps even largely, a collective, network phenomenon

382 citations


Journal ArticleDOI
TL;DR: Within the evolutionary computation (EC) literature, this is known as co-evolution and within this paradigm, expert game-playing strategies have been evolved without the need for human expertise.
Abstract: Games provide competitive, dynamic environments that make ideal test beds for computational intelligence theories, architectures, and algorithms. Natural evolution can be considered to be a game in which the rewards for an organism that plays a good game of life are the propagation of its genetic material to its successors and its continued survival. In natural evolution, the fitness of an individual is defined with respect to its competitors and collaborators, as well as to the environment. Within the evolutionary computation (EC) literature, this is known as co-evolution and within this paradigm, expert game-playing strategies have been evolved without the need for human expertise.

133 citations


Journal ArticleDOI
TL;DR: A unified framework called pseudo-label fuzzy support vector machine (PLFSVM), an extended version of SVM, takes into account the fuzzy nature of some training samples during its training to perform content-based image retrieval.
Abstract: Conventional relevance feedback in content-based image retrieval (CBIR) systems uses only the labeled images for learning. Image labeling, however, is a time-consuming task and users are often unwilling to label too many images during the feedback process. This gives rise to the small sample problem where learning from a small number of training samples restricts the retrieval performance. To address this problem, we propose a technique based on the concept of pseudo-labeling in order to enlarge the training data set. As the name implies, a pseudo-labeled image is an image not labeled explicitly by the users, but estimated using a fuzzy rule. Therefore, it contains a certain degree of uncertainty or fuzziness in its class information. Fuzzy support vector machine (FSVM), an extended version of SVM, takes into account the fuzzy nature of some training samples during its training. In order to exploit the advantages of pseudo-labeling, active learning and the structure of FSVM, we develop a unified framework called pseudo-label fuzzy support vector machine (PLFSVM) to perform content-based image retrieval. Experimental results based on a database of 10,000 images demonstrate the effectiveness of the proposed method

81 citations


Journal ArticleDOI
TL;DR: In this article, a salting route optimization (SRO) system that combines evolutionary algorithms with the neXt generation Road Weather Information System (XRWIS) is introduced, which means that salting routes optimization can be done at a level previously not possible.
Abstract: Highway authorities in marginal winter climates are responsible for the precautionary gritting/salting of the road network in order to prevent frozen roads. For efficient and effective road maintenance, accurate road surface temperature prediction is required. However, this information is useless if an effective means of utilizing this information is unavailable. This is where gritting route optimization plays a crucial role. The decision whether to grit the road network at marginal nights is a difficult problem. The consequences of making a wrong decision are serious, as untreated roads are a major hazard. However, if grit/salt is spread when it is not actually required, there are unnecessary financial and environmental costs. The goal here is to minimize the financial and environmental costs while ensuring roads that need treatment will. In this article, a salting route optimization (SRO) system that combines evolutionary algorithms with the neXt generation Road Weather Information System (XRWIS) is introduced. The synergy of these methodologies means that salting route optimization can be done at a level previously not possible.

77 citations


Journal ArticleDOI
TL;DR: This paper visualize the struc- ture and the evolution of the compu- tational intelligence (CI) field and analyses the way in which the CI field is divided into subfields, providing insight into the characteristics of each subfield and into the relations between the subfields.
Abstract: In this paper, we visualize the structure and the evolution of the computational intelligence (CI) field. Based on our visualizations, we analyze the way in which the CI field is divided into several subfields. The visualizations provide insight into the characteristics of each subfield and into the relations between the subfields. By comparing two visualizations, one based on data from 2002 and one based on data from 2006, we examine how the CI field has evolved over the last years. A quantitative analysis of the data further identifies a number of emerging areas within the CI field.

71 citations


Journal ArticleDOI
TL;DR: This article surveys evolvable hardware with emphasis on some of the latest developments, many of which deliver performance exceeding traditional methods.
Abstract: Evolvable hardware lies at the intersection of evolutionary computation and physical design. Through the use of evolutionary computation methods, the field seeks to develop a variety of technologies that enable automatic design, adaptation, and reconfiguration of electrical and mechanical hardware systems in ways that outperform conventional techniques. This article surveys evolvable hardware with emphasis on some of the latest developments, many of which deliver performance exceeding traditional methods. As such, the field of evolvable hardware is just now starting to emerge from the research laboratory and into mainstream hardware applications.

68 citations


Journal ArticleDOI
TL;DR: This paper analyzes the way in which the CI field is divided into several subfields and provides insight into the characteristics of each subfield and into the relations between the subfields.
Abstract: In this paper, we visualize the structure and the evolution of the computational intelligence (CI) field. Based on our visualizations, we analyze the way in which the CI field is divided into several subfields. The visualizations provide insight into the characteristics of each subfield and into the relations between the subfields. By comparing two visualizations, one based on data from 2002 and one based on data from 2006, we examine how the CI field has evolved over the last years. A quantitative analysis of the data further identifies a number of emerging areas within the CI field. The data that we use consist of the abstracts of the papers presented at the IEEE World Congress on Computational Intelligence (WCCI) in 2002 and 2006. Using a fully automatic procedure, so-called concept maps are constructed from the data. These maps visualize the associations between the main concepts in the CI field. Our analysis of the structure and the evolution of the CI field are largely based on the constructed concept maps

63 citations



Journal ArticleDOI
TL;DR: Why autonomous development is necessary according to a concept called task muddiness is discussed and results for a series of research issues are introduced, including the new paradigm for autonomous development, mental architectures, developmental algorithm, a refined classification of types of machine learning, spatial complexity and time complexity.
Abstract: Artificial neural networks can model cortical local learning and signal processing, but they are not the brain, neither are many special purpose systems to which they contribute. Autonomous mental development models all or part of the brain (or the central nervous system) and how it develops and learns autonomously from infancy to adulthood. Like neural network research, such modeling aims to be biologically plausible. This paper discusses why autonomous development is necessary according to a concept called task muddiness. Then it introduces results for a series of research issues, including the new paradigm for autonomous development, mental architectures, developmental algorithm, a refined classification of types of machine learning, spatial complexity and time complexity. Finally, the paper presents some experimental results for applications, including: vision-guided navigation, object finding, object-based attention (eye-pan), and attention-guided pre-reaching, tour tasks that infants learn to perform early but very perceptually challenging for robots

Journal ArticleDOI
TL;DR: The textbook “Soft Computing and Intelligent Systems Design, Theory, Tools and Applications” by Fakhreddine Karray and Clarence W. de Silva represents a comprehensive and cohesive treatment of the state-of-theart consortium of soft-computing methodologies and their potential integration, from both the analytical and the practical perspectives.
Abstract: The textbook “Soft Computing and Intelligent Systems Design, Theory, Tools and Applications” by Fakhreddine Karray and Clarence W. de Silva represents a comprehensive and cohesive treatment of the state-of-theart consortium of soft-computing methodologies and their potential integration, from both the analytical and the practical perspectives. The textbook thoroughly details the technical aspects of its topics to better serve practicing professionals, yet it does not sacrifice clarity of presentation and simplicity in style, which makes it also appealing for students and novice researchers in the textbook’s related fields. The large number of illustrative examples, end-ofchapter problems, and solved case studies in various engineering applications make the textbook an excellent choice for a wide range of courses in interdisciplinary engineering fields. The content structure of the book suits courses in areas such as fuzzy logic, neural networks, evolutionary computing, machine intelligence, and intelligent control. The textbook is organized into four main parts. The first part presents soft computing and its applications including intelligent control. The second part deals with the various types of connectionist modeling techniques and their applications. Part three discusses evolutionary computing algorithms and their synergistic integration with neural networks. Part four demonstrates the use of the a priori-discussed techniques through a number of worked case studies taken from real-world applications in various engineering disciplines. Chapters 1 through 3 comprise Part 1. Chapter 1 elegantly introduces machine intelligence and outlines tools of soft computing and their merits for the design of a wide range of intelligent systems. Chapter 2 uses realworld examples to motivate the use of fuzzy set theory before it tackles the fundamental aspects and the theoretical background of the topic. The discussed concepts are illustrated with a number of examples to help the reader grasp the theory behind the discussed concepts. Chapter 3 is devoted to discussing the major aspects of fuzzy logic control. It also explores different strategies in designing fuzzy logic controllers while enumerating their properties and popular applications. Part 2 on neural networks and their integration into dynamic fuzzy models covers Chapters 4 through 7. Chapter 4 provides a general introduction to artificial neural networks. Chapter 5 details the major classes of neural networks with a particular emphasis on multilayer perceptrons, radial basis function networks, Kohonen’s self-organizing networks, and Hopfield networks. The chapter also presents an extensive review on the applications of these types of neural networks in a wide spectrum of industrial applications. Chapter 6 introduces the concept of dynamic neural networks and illustrates their use in dynamic processes and chaos-time series prediction. Chapter 7 bridges Parts 1 and 2 by revealing different cutting-edge methodologies for the synergistic integration of fuzzy logic and neural networks. Part 3 consists of Chapter 8, which starts with an introduction on evolutionary computing. It then lays the fundamentals of genetic algorithms including the schema theorem. The rest of the chapter is devoted to reviewing selected techniques for integrating genetic algorithms with neural networks and fuzzy logic. The chapter is concluded with a set of


Journal ArticleDOI
TL;DR: The primary focus of this presentation will be on the usage of so-called tiny-GAs, an evolutionary procedure with low efforts, i.e. small population size, little number of generations, and a simple fitness, which is suitable for solving highly complex optimization tasks.
Abstract: The expedience of today's image-processing applications is no longer based on the performance of a single algorithm alone. These systems appear to be complex frameworks with a lot of sub-tasks that are solved by specific algorithms, adaptation procedures, data handling, scheduling, and parameter choices. The venture of using computational intelligence (CI) in such a context, thus, is not a matter of a single approach. Among the great choice of techniques to inject CI in an image-processing framework, the primary focus of this presentation will be on the usage of so-called tiny-GAs. This stands for an evolutionary procedure with low efforts, i.e. small population size (like 10 individuals), little number of generations, and a simple fitness. Obviously, this is not suitable for solving highly complex optimization tasks, but the primary interest here is not the best individual's fitness, but the fortune of the algorithm and its population, which has just escaped the Monte-Carlo domain after random initialization. That this approach can work in practice will be demonstrated by means of selected image-processing applications, especially in the context of linear regression and line fitting; evolutionary post processing of various clustering results, in order to select a most suitable one by similarity; and classification by the fitness values obtained after a few generations

Journal ArticleDOI
TL;DR: This paper describes an implementation of a multizone temperature experiment, and shows that the use of multiple agents defines a distributed controller that can equilibrate the temperatures in the zones in spite of interzone, ambient, and network effects.
Abstract: Models from behavioral ecology, specifically foraging theory, are used to describe the decisions an animal forager must make in order to maximize its rate of energy gain and thereby improve its survival probability. Using a bioinspired methodology, we view an animal as a software agent, the foraging landscape as a spatial layout of temperature zones, and nutrients as errors between the desired and actual temperatures in the zones. Then, using foraging theory, we define a decision strategy for the agent that has an objective of reducing the temperature errors in order to track a desired temperature. We describe an implementation of a multizone temperature experiment, and show that the use of multiple agents defines a distributed controller that can equilibrate the temperatures in the zones in spite of interzone, ambient, and network effects. We discuss relations to ideas from theoretical ecology, and identify a number of promising research directions. It is our hope that the results of this paper will motivate other research on bioinspired methods based on behavioral ecology

Journal ArticleDOI
TL;DR: Nils Aall Barricelli was one of the pioneers in evolutionary computation and publication of "Esempi Numerici di processi di evoluzione", a study in what would now be called artificial life, in 1954 is perhaps the earliest published record of an evolutionary simulation.
Abstract: Nils Aall Barricelli was one of the pioneers in evolutionary computation. His publication of "Esempi Numerici di processi di evoluzione", a study in what would now be called artificial life, in the journal Methodos, in 1954 is perhaps the earliest published record of an evolutionary simulation. The paper was republished in 1957 in English, and detailed the results of programs that were run at the Institute for Advanced Study in Princeton, NJ, in 1953.



Journal ArticleDOI
TL;DR: The evolutionary multi-objective optimization (EMOO) repository has become much more than the simple list of bibliographic references that originated it and has become a valuable source for students and researchers interested in this area.
Abstract: This article briefly describes the evolutionary multi-objective optimization (EMOO) repository, which has become much more than the simple list of bibliographic references that originated it. In its current state, the EMOO repository contains many Web resources, including Ph.D. theses, software, contact information of EMOO researchers and information about EMOO-related events. Such information has become a valuable source for students and researchers interested in this area.

Journal ArticleDOI
TL;DR: Developmental psychology is ready to blossom into a modern science that focuses on causal mechanistic explanations of development rather than just describing and classifying the skills that children show at different ages.
Abstract: Developmental psychology is ready to blossom into a modern science that focuses on causal mechanistic explanations of development rather than just describing and classifying the skills that children show at different ages. Computational models of cognitive development are formal systems that track the changes in information processing taking place as a behavior is acquired. Models are generally implemented as psychologically constrained computer simulations that learn tasks such as reasoning, categorization, and language. Their principal use is as tools for exploring mechanisms of transition (development) from one level of competence to the next during the course of cognitive development. They have been used to probe questions such as the extent of 'pre-programmed' or innate knowledge that exists in the infant mind, and how the sophistication of reasoning can increase with age and experience

Journal ArticleDOI
TL;DR: The research group builds robots that learn in the same type of supportive environment that human children have and develop skills incrementally through their interactions and sits at the intersection of the fields of social robotics and autonomous mental development.
Abstract: Most robots are designed to operate in environments that are either highly constrained (as is the case in an assembly line) or extremely hazardous (such as the surface of Mars). Machine learning has been an effective tool in both of these environments by augmenting the flexibility and reliability of robotic systems, but this is often a very difficult problem because the complexity of learning in the real world introduces very high dimensional state spaces and applies severe penalties for mistakes. Human children are raised in environments that are just as complex (or even more so) than those typically studied in robot learning scenarios. However, the presence of parents and other caregivers radically changes the type of learning that is possible. Consciously and unconsciously, adults tailor their action and the environment to the child. They draw attention to important aspects of a task, help in identifying the cause of errors and generally tailor the task to the child's capabilities. Our research group builds robots that learn in the same type of supportive environment that human children have and develop skills incrementally through their interactions. Our robots interact socially with human adults using the same natural conventions that a human child would use. Our work sits at the intersection of the fields of social robotics (Fong et al., 2003; Breazeal and Scawellan, 2002) and autonomous mental development (Weng et al., 2000). Together, these two fields offer the vision of a machine that can learn incrementally, directly from humans, in the same ways that humans learn from each other. In this article, we introduce some of the challenges, goals, and applications of this research

Journal ArticleDOI
Xin Yao1
TL;DR: This article summarises one such MSc program in natural computation at the University of Birmingham, UK.
Abstract: Natural computation is the study of computational systems that use ideas and get inspirations from natural systems, including biological, physical, chemical, economical and social systems. It covers many active research fields, such as evolutionary computation, neural computation, molecular computation, quantum computation, ecological computation, etc. It has made tremendous progress in academic research and real-world applications. Many university departments have been offering individual modules/courses in one form or another on related topics. However, few universities have been offering an entire postgraduate program in natural computation. This article summarises one such MSc program in natural computation at the University of Birmingham, UK.

Journal ArticleDOI
TL;DR: The final author version and the galley proof are versions of the publication after peer review and the final published version features the final layout of the paper including the volume, issue and page numbers.
Abstract: • A submitted manuscript is the author's version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website. • The final author version and the galley proof are versions of the publication after peer review. • The final published version features the final layout of the paper including the volume, issue and page numbers.

Journal ArticleDOI
TL;DR: The Evolutionary Computation Benchmarking Repository (EvoCoBR) has been designed and put into operation in a beta version and trial phase, and the architecture enables the entire evolutionary computation community to contribute and own the Web-based archive.
Abstract: Evolutionary computation has been used with great success for the solution of hard optimization problems. Theoretical analysis, although important in its own right, e.g. for understanding underlying phenomena and characteristics of evolutionary search, can only provide upper and/or lower bounds of performance estimation of evolutionary algorithms for hard optimization problems. In practice, empirical analysis is the most important means to assess and compare the performance of algorithms. In order to facilitate this fair and transparent comparison, the Evolutionary Computation Benchmarking Repository (EvoCoBR) by M. Roberts et al. (2006) has been designed and put into operation in a beta version and trial phase. The aim is to create a central Web-based repository for storing detailed benchmark problem descriptions. However, with EvoCoBR we want to go one step further and archive, along with the problem description, a list of references to previously achieved results and the best result so far. This enables researchers to more easily see how their results compare to results in the literature. EvoCoBR will also invite researchers to submit and archive the programs that produced those results. EvoCcBR's architecture enables the entire evolutionary computation community to contribute and own the Web-based archive. Its contents will be submitted by researchers and practitioners, and openly accessible by all. In other words, the EvoCoBR design defines the framework that needs to be filled by the evolutionary computation community for the evolutionary computation community

Journal ArticleDOI
TL;DR: Methods for fast search through finite-state machines, Bayesian solutions for modeling and classification of speech, and a discriminative training approach for minimizing errors in large vocabulary continuous speech recognition are focused on.
Abstract: Recent developments in research on humanoid robots and interactive agents have highlighted the importance of and expectation on automatic speech recognition (ASR) as a means of endowing such an agent with the ability to communicate via speech. This article describes some of the approaches pursued at NTT Communication Science Laboratories (NTT-CSL) for dealing with such challenges in ASR. In particular, we focus on methods for fast search through finite-state machines, Bayesian solutions for modeling and classification of speech, and a discriminative training approach for minimizing errors in large vocabulary continuous speech recognition

Journal ArticleDOI
TL;DR: The editors, who are well-recognized researchers in the design of MOEAs, have done an excellent work in compiling a representative set of applications that is clearly state-of-the-art.
Abstract: Solving real-world problems involves conflicting objectives; this implies that an ideal point that is optimum under all objectives at the same time does not exist. The definition of what the optimum is and how to find it, under these conflicting objectives, has led to the development of multiobjective (MO) optimization [1-2]. The editors, who are well-recognized researchers in the design of MOEAs, have done an excellent work in compiling a representative set of applications that is clearly state-of-the-art. The authors of each chapter contribute their point of view about how to apply MOEAs to solve specific problems, which enriches the technical content of the book. The book is about applications that resist being solved by traditional methods. The book is conveniently organized in Chapter 1, and the remaining 29 chapters are divided into four parts. Chapter 1 gives a brief introduction to the field of MOEA and sets up the basic concepts needed to understand the rest of the book, along with a brief description of each chapter. Applications in engineering are dealt with in Chapters 2–13 (Part I). Scientific applications are the concern in Chapters 14–19 (Part II). Chapters 20–24 (Part III) deal with industrial applications. Chapters 25–30 (Part IV) deal with miscellaneous applications. Each chapter, in these parts, is organized as follows.


Journal ArticleDOI
TL;DR: This article provides a brief introduction to hierarchical discriminant regression and lobe component analysis, two algorithms usable in components of the mental architecture of an AMD system.
Abstract: This article provides a brief introduction to hierarchical discriminant regression and lobe component analysis, two algorithms usable in components of the mental architecture of an AMD system. The processes that guide the development of the "mind" of an autonomously-learning robot must be given within the developmental program. This is what must drive the learning of the robot's eventual behaviors and skills, which emerges in response to the robot's interactions with the environment. A biologically sophisticated version of a developmental program, where behaviors and skills autonomously emerge in the robot with a human level of complexity, has not yet been written. But there has been progress in creating developmental "building blocks." These are designed to be used in components of the developmental program. These algorithms must be specially designed to handle many of the difficult constraints of AMD. In this article, we introduce two such algorithms, briefly outline their purpose and key features, and describe how to use sample versions available on the El lab's Web site

Journal ArticleDOI
TL;DR: A new family of neural network architectures that have a basis inSelf-organization, yet are somewhat free from many of the constraints typical of other well-known self-organizing architectures are explored.
Abstract: In this article, we explore a new family of neural network architectures that have a basis in self-organization, yet are somewhat free from many of the constraints typical of other well-known self-organizing architectures. Within this family, the basic processing unit is known as the self-organizing tree map (SOTM). We will look at how this model has evolved since its inception in 1995, how it has inspired new models, and how it is being applied to complex multimedia research problems in digital asset management and microbiological image analysis