Dimitris E. Koulouriotis

Comparing simulated annealing and genetic algorithm in learning FCM

Fuzzy Cognitive Map (FCM) is a directed graph, which shows the relations between essential components in complex systems. It is a very convenient, simple, and powerful tool, which is used in numerous areas of application. Experts who are familiar with the system components and their relations can generate a related FCM. There is a big gap when human experts cannot produce FCM or even there is no expert to produce the related FCM. Therefore, a new mechanism must be used to bridge this gap. In this paper, a novel learning method is proposed to construct FCM by using some metaheuristic methods such as genetic algorithm (GA) and simulated annealing (SA). The proposed method not only is able to construct FCM graph topology but also is able to extract the weight of the edges from input historical data. The efficiency of the proposed method is shown via comparison of its results of some numerical examples with those of some other methods.

Fuzzy Cognitive Maps for Pattern Recognition Applications

A first attempt to incorporate Fuzzy Cognitive Maps (FCMs), in pattern classification applications is performed in this paper. Fuzzy Cognitive Maps, as an illustrative causative representation of modeling and manipulation of complex systems, can be used to model the behavior of any system. By transforming a pattern classification problem into a problem of discovering the way the sets of patterns interact with each other and with the classes that they belong to, we could describe the problem in terms of Fuzzy Cognitive Maps. More precisely, some FCM architectures are introduced and studied with respect to their pattern recognition abilities. An efficient novel hybrid classifier is proposed as an alternative classification structure, which exploits both neural networks and FCMs to ensure improved classification capabilities. Appropriate experiments with four well-known benchmark classification problems and a typical computer vision application establish the usefulness of the Fuzzy Cognitive Maps, in a pattern recognition research field. Moreover, the present paper introduces the use of more flexible FCMs by incorporating nodes with adaptively adjusted activation functions. This advanced feature gives more degrees of freedom in the FCM structure to learn and store knowledge, as needed in pattern recognition tasks.

Development of dynamic cognitive networks as complex systems approximators: validation in financial time series

Dynamic cognitive networks (DCNs) define a novel approach to functionalize cognitive mapping and complex systems analysis, which were recently supported by fuzzy cognitive maps (FCMs). The modeling and inference limitations met in FCMs, especially in situations with strong nonlinearity and temporal phenomena, pushed towards DCNs; their theoretical framework is scheduled to confront the preceding weaknesses and offer wider possibilities in causal structures management. Trying to contribute to the enhancement of DCNs, at first, systemic and environmental metaphors are introduced with practical mathematical formalisms and generalized nomenclature. Nonlinear and asymmetric cause-effect relationships, decaying mechanisms, inertial forces, diminishing effects and biases formulate a powerful set of adaptive characteristics that strengthen the operational behavior of DCNs. Second, the strategic reorientation of DCNs is attempted as generalized approximation tools. This new strategic option is verified not only in classical function approximation tests, but also in the challenging area of securities markets. The platform of evaluation of DCNs involves comparisons with a linear multiple regression model, a feed-forward neural network trained with both back-propagation and evolution strategies, a radial basis function network, and an adaptive network-based fuzzy inference system (ANFIS). Through the experiments for short-term stock price predictions, multiple issues are analyzed not only about the role of diverse DCN parameters, but also about the given problem of financial markets modeling and forecasting. Simulations distinguish DCNs as a strong methodology with noticeable adaptability in complicated patterns and broad generalization capabilities while, at the same time, the all-embracing outcomes support previous findings of partially random walk phenomena in short-term stock market forecasting attempts.

Novel moment invariants for improved classification performance in computer vision applications

A novel set of moment invariants based on the Krawtchouk moments are introduced in this paper. These moment invariants are computed over a finite number of image intensity slices, extracted by applying an innovative image representation scheme, the image slice representation (ISR) method. Based on this technique an image is decomposed to a several non-overlapped intensity slices, which can be considered as binary slices of certain intensity. This image representation gives the advantage to accelerate the computation of images moments since the image can be described in a number of homogenous rectangular blocks, which permits the simplification of the computation formulas. The moments computed over the extracted slices seem to be more efficient than the corresponding moments of the same order that describe the whole image, in recognizing the pattern under processing. The proposed moment invariants are exhaustively tested in several well known computer vision datasets, regarding their rotation, scaling and translation (RST) invariant recognition performance, by resulting to remarkable outcomes.

Reinforcement learning and evolutionary algorithms for non-stationary multi-armed bandit problems

Abstract Multi-armed bandit tasks have been extensively used to model the problem of balancing exploitation and exploration. A most challenging variant of the MABP is the non-stationary bandit problem where the agent is faced with the increased complexity of detecting changes in its environment. In this paper we examine a non-stationary, discrete-time, finite horizon bandit problem with a finite number of arms and Gaussian rewards. A family of important ad hoc methods exists that are suitable for non-stationary bandit tasks. These learning algorithms that offer intuition-based solutions to the exploitation–exploration trade-off have the advantage of not relying on strong theoretical assumptions while in the same time can be fine-tuned in order to produce near-optimal results. An entirely different approach to the non-stationary multi-armed bandit problem presents itself in the face of evolutionary algorithms. We present an evolutionary algorithm that was implemented to solve the non-stationary bandit problem along with ad hoc solution algorithms, namely action-value methods with e-greedy and softmax action selection rules, the probability matching method and finally the adaptive pursuit method. A number of simulation-based experiments was conducted and based on the numerical results that we obtained we discuss the methods’ performances.

Moment-based local binary patterns: A novel descriptor for invariant pattern recognition applications

A novel descriptor able to improve the classification capabilities of a typical pattern recognition system is proposed in this paper. The introduced descriptor is derived by incorporating two efficient region descriptors, namely image moments and local binary patterns (LBP), commonly used in pattern recognition applications, in the last decades. The main idea behind this novel feature extraction methodology is the need of improved recognition capabilities, a goal achieved by the combinative use of these descriptors. This collaboration aims to make use of the major advantages each one presents, by simultaneously complementing each other, in order to elevate their weak points. In this way, the useful properties of the moments and moment invariants regarding their robustness to the noise presence, their global information coding mechanism and their invariant behaviour under scaling, translation and rotation conditions, along with the local nature of the LBP, are combined in a single concrete methodology. As a result a novel descriptor invariant to common geometric transformations of the described object, capable to encode its local characteristics, is formed and its classification capabilities are investigated through massive experimental scenarios. The experiments have shown the superiority of the introduced descriptor over the moment invariants, the LBP operator and other well-known from the literature descriptors such as HOG, HOG-LBP and LBP-HF.

A unified methodology for the efficient computation of discrete orthogonal image moments

A novel methodology is proposed in this paper to accelerate the computation of discrete orthogonal image moments. The computation scheme is mainly based on a new image representation method, the image slice representation (ISR) method, according to which an image can be expressed as the outcome of an appropriate combination of several non-overlapped intensity slices. This image representation decomposes an image into a number of binary slices of the same size whose pixels come in two intensities, black or any other gray-level value. Therefore the image block representation can be effectively applied to describe the image in a more compact way. Once the image is partitioned into intensity blocks, the computation of the image moments can be accelerated, as the moments can be computed by using decoupled computation forms. The proposed algorithm constitutes a unified methodology that can be applied to any discrete moment family in the same way and produces similar promising results, as has been concluded through a detailed experimental investigation.

Anamorphosis of fuzzy cognitive maps for operation in ambiguous and multi-stimulus real world environments

FCMs were presented in the middle eighties and extended through studies, which enhanced the basic principles and proposed domains of successful application. The special scientific interest about FCMs is due to their favorable characteristics, in conjunction with the representation ability and the inference mechanism. Indeed, FCMs embrace flexibility, functionality and simplicity, while simultaneously they maintain the advantageous features of typical fuzzy systems. Although the core concepts of FCM theory have been clarified, robustness of the functional framework is still under question, as various aspects have not been explicitly described and in many cases, the current practice seems to restrict efficiency and credibility. The present work aims to improve the FCM background by conducting a thorough analysis of the basic principles, exposing modifications of the operating framework and, mainly, introducing an innovative inference procedure that manages more reasonably multi-stimulus situations, which constitute a common phenomenon in the real world.

Towards Hebbian learning of Fuzzy Cognitive Maps in pattern classification problems

A detailed comparative analysis of the Hebbian-like learning algorithms applied to train Fuzzy Cognitive Maps (FCMs) operating as pattern classifiers, is presented in this paper. These algorithms aim to find appropriate weights between the concepts of the FCM classifier so it equilibrates to a desired state (class mapping). For these purposes, six different types of Hebbian learning algorithms from the literature have been selected and studied in this work. Along with the theoretical description of these algorithms and the analysis of their performance in classifying known patterns, a sensitivity analysis of the applied classification scheme, regarding some configuration parameters have taken place. It is worth noting that the algorithms are studied in a comparative fashion, under common configurations for several benchmark pattern classification datasets, by resulting to useful conclusions about their training capabilities.

A fuzzy cognitive map-based stock market model: synthesis, analysis and experimental results

The expansion of advanced modeling tools, such as neural, evolutionary, fuzzy and hybrid systems, has led to a systematic attempt for their applicability in the challenging stock market field. Today, the ensuing results are admittedly far better than those accomplished by models based on linear or typical nonlinear mathematical approximators; yet, the related trading risk remains at significantly high levels. In quest of innovative approaches, one interesting research direction appears to be the complete analysis and exploitation of various interrelated quantitative and mostly qualitative agents affecting stock market behavior. Based on this criterion, fuzzy cognitive maps (FCMs) constitute a powerful modeling tool for the development of a stock market forecasting system as they are structured as networks of cause-effect relationships between diverse factors. The subject of this study is aligned with the aforementioned remark; firstly, the recognition of crucial stock market, business and economic agents is attempted, secondly an FCM-based stock market model is designed, and ultimately the feasibility and effectiveness of the real world application is evaluated.