Recent Advances in Intelligent Paradigms and Applications

The integration of different learning and adaptation techniques to overcome individual limitations and to achieve synergetic effects through the hybridization or fusion of these techniques has, in recent years, contributed to a large number of new intelligent system designs. Computational intelligence is an innovative framework for constructing intelligent hybrid architectures involving Neural Networks (NN), Fuzzy Inference Systems (FIS), Probabilistic Reasoning (PR) and derivative free optimization techniques such as Evolutionary Computation (EC). Most of these hybridization approaches, however, follow an ad hoc design methodology, justified by success in certain application domains. Due to the lack of a common framework it often remains difficult to compare the various hybrid systems conceptually and to evaluate their performance comparatively. This chapter introduces the different generic architectures for integrating intelligent systems. The designing aspects and perspectives of different hybrid archirectures like NN-FIS, EC-FIS, EC-NN, FIS-PR and NN-FIS-EC systems are presented. Some conclusions are also provided towards the end.

This chapter presents an Hybrid Architecture based on Representations, Perception and Intelligent Control (HARPIC). It includes reactive and deliberative behaviors, which we have developed to confer autonomy to unmanned robotics systems. Two main features characterize our work: on the one hand the ability for the robot to control its own autonomy, and on the other hand the capacity to evolve and to learn.

We propose a new approach for medical diagnosis by employing intuitionistic fuzzy sets (cf. Atanassov [1], [2]) which because of additional degree of freedom in comparison with fuzzy sets (Zadeh [14]), can be viewed as their generalization Employing intuitionistic fuzzy sets, we can simply and adequately express a hesitation concerning the objects considered — both patients and illnesses. Solution is obtained by looking for the smallest distance (cf. Szmidt and Kacprzyk [8], [11]) between symptoms that are characteristic for a patient and symptoms describing illnesses considered. We point out advantages of this new technique over the method proposed by De, Biswas and Roy [4] where intuitionistic fuzzy sets were also applied but the max-min-max composition of intuitionistic fuzzy relations was used instead of taking into account all, unchanged symptom values as proposed in this article.

Nowadays, people start to accept fuzzy rule-based systems as flexible and convenient tools to solve a myriad of ill-defined but otherwise (for humans) straightforward tasks such as controlling fluid levels in a reactor, automatical lens focussing in cameras and adjusting an aircraft’s navigation to the change of winds and so on. Contrary to the intuition often seen as the feeding ground of fuzzy rule-based systems—namely, that they realize an extension of the Modus Ponens (MP) rule of inference to an environment with more than two truth-values—most actual applications rely at the base level on common interpolation techniques or similarity assessments to simulate the process of “calculating with words” perceived at the user level. It is doubtful whether these somewhat opportunistic approaches will perform well when more challenging requirements (e.g. aspects of logical consistency; incorporation of varying facets of uncertainty) are imposed in order to implement a successful artificial reasoning unit. Therefore, in this paper, starting from the notion of a fuzzy restriction (i.e. the basic building block of our rule-based system) we list some elementary consistency requirements that a fuzzy inference system should satisfy. Subsequently we describe a reasoning methodology based on a measure of fulfilment of the antecedent clause of an if-then rule. Inclusion-based approximate reasoning, as we coined it in [7], outperforms the traditional scheme based on the Compositional Rule of Inference (CRI) in terms of both complexity and of logical soundness. In terms of semantics it also offers a better solution to the implementation of analogical reasoning than similarity measures are able to do.

Job-shop scheduling is a difficult problem, both theoretically and practically. The theoretical problems stem from the search for optimal schedules subject to a limited number of constraints, while the complexity of practical problems is due to the number and variety of constraints that are not rigid in the actual situations. Actually, the job-shop scheduling problem is a combinatorial optimization of considerable industrial importance. Although this problem has often been investigated, very little of this research is concerned with the uncertainty characterized by the imprecision in problem variables. In this paper, we investigate a fuzzy approach to the job-shop scheduling problem based on imprecise processing times. We first use triangular fuzzy numbers to represent imprecise processing times, and then construct a fuzzy job-shop scheduling model to solve the problem. Our work intends to extend the crisp job-shop scheduling problem, in the case of imprecise processing times, into a more generalized problem that would be useful in practical situations.

A natural approach to designing an intelligent system is to incorporate expert knowledge into this system. One of the main approaches to translating this knowledge into computer-understandable terms is the approach of fuzzy logic. It has led to many successful applications, but in several aspects, the resulting computer representation is somewhat different from the original expert meaning. Two related approaches have been used to make fuzzy logic more adequate in representing expert reasoning: granularity and higher-order approaches. Each approach is successful in some applications where the other approach did not succeed so well; it is therefore desirable to combine these two approaches. This idea of combining the two approaches is very natural, but so far, it has led to few successful practical applications. In this chapter, we provide results aimed at finding a better (ideally optimal) way of combining these approaches.

The Standard BackPropagation (SBP) algorithm is the most widely known and used learning method for training neural networks. Unfortunately, SBP suffers from several problems such as sensitivity to the initial conditions and very slow convergence. Here we describe how we used Genetic Programming, a search algorithm inspired by Darwinian evolution, to discover new supervised learning algorithms for neural networks which can overcome some of these problems. Comparing our new algorithms with SBP on different problems we show that these are faster, are more stable and have greater feature extracting capabilities.

Modern processes in industry are characterized by nonlinear and time-varying behavior. Nonlinear system identification is becoming an important tool which can be used to improve control performance and achieve robust fault-tolerant behavior. Among the different nonlinear identification techniques, methods based on neuro-fuzzy models are gradually becoming established not only in the academia but also in industrial applications. Neuro-fuzzy modeling can be regarded as a gray-box technique on the boundary between neural networks and qualitative fuzzy models. The tools for building neuro-fuzzy models are based on combinations of algorithms from the fields of neural networks, pattern recognition and regression analysis. This chapter addresses the use of neuro-fuzzy models in system identification.

In this chapter, a new genetic algorithm is proposed for packing rectangular cargos of different sizes into a given area in a two dimensional framework. A novel penalty function method is proposed for checking the solution strings that have violated the loading area’s aspect constraints. This method penalizes the solution strings based on the extent by which they violate the length and breadth constraints. The proposed genetic algorithm is compared with the heuristic method of Gehring et.al. and the genetic algorithm of Hwang et. al. Results indicate that the proposed method is superior in terms of packing efficiency and solution time.

This chapter compares the performance of six evolutionary algorithms, three sequential and three parallel, for solving combinatorial optimization problems. In particular, a generational, a steady-state, a cellular genetic algorithm, and their distributed versions were applied to the maximum cut problem, the error correcting code design problem, and the minimum tardy task problem. The algorithms were tested on a total of seven problem instances. The results obtained in this chapter are better than the ones previously reported in the literature in all cases except for one problem instance. The high quality results were achieved although no problem-specific changes of the evolutionary algorithms were made other than in the fitness function. Just the intrinsic search features of each class of algorithms proved to be powerful enough to solve a given problem instance. Some of the sequential, and almost every parallel algorithm, yielded fast and accurate results, although they sampled only a tiny fraction of the search space.

In this chapter we discuss how we have generated nonverbal output through an embodied agent, based on a user’s actions in an ITS. Our project has been principally concerned with maintaining an emotional state for a virtual character. Presented herein is the basic emotional model we used for the character’s internal emotion management through qualitative reasoning. We give an overview of the agent’s environment and describe the role the agent is designed to play, using our own system as a reference; next, we outline the agent’s internal architecture. In conclusion, we discuss the inputs taken by the system and how these are treated to modify the emotional model of the agent.

Finding suitable jobs for US Navy sailors from time to time is an important and ever-changing process. An Intelligent Distribution Agent and particularly its constraint satisfaction module take up the challenge to automate the process. The constraint satisfaction module’s main task is to assign sailors to new jobs in order to maximize Navy and sailor happiness. We present various neural network techniques combined with several statistical criteria to optimize the module’s performance and to make decisions in general. The data was taken from Navy databases and from surveys of Navy experts. Such indeterminate subjective component makes the optimization of the constraint satisfaction a very sophisticated task. Single-Layer Perceptron with logistic regression, Multilayer Perceptron with different structures and algorithms and Support Vector Machine with Adatron algorithm are presented for achieving best performance. Multilayer Perceptron neural network and Support Vector Machine with Adatron algorithm produced highly accurate classification and encouraging prediction.