Search

In this paper we propose a procedure to enable the training of several independent Multilayer Perceptron Neural Networks with a different number of neurons and activation functions in parallel (ParallelMLPs) by exploring the principle of locality and parallelization capabilities of modern CPUs and GPUs. The core idea of this technique is to represent several sub-networks as a single large network and use a Modified Matrix Multiplication that replaces an ordinal matrix multiplication with two simple matrix operations that allow separate and independent paths for gradient flowing. We have assessed our algorithm in simulated datasets varying the number of samples, features and batches using 10,000 different models as well as in the MNIST dataset. We achieved a training speedup from 1 to 4 orders of magnitude if compared to the sequential approach. The code is available online. [ABSTRACT FROM AUTHOR]

Spiking neural networks (SNNs) aim to replicate energy efficiency, learning speed and temporal processing of biological brains. However, accuracy and learning speed of such networks is still behind reinforcement learning (RL) models based on traditional neural models. This work combines a pre-trained binary convolutional neural network with an SNN trained online through reward-modulated STDP in order to leverage advantages of both models. The spiking network is an extension of its previous version, with improvements in architecture and dynamics to address a more challenging task. We focus on extensive experimental evaluation of the proposed model with optimized state-of-the-art baselines, namely proximal policy optimization (PPO) and deep Q network (DQN). The models are compared on a grid-world environment with high dimensional observations, consisting of RGB images with up to 256 × 256 pixels. The experimental results show that the proposed architecture can be a competitive alternative to deep reinforcement learning (DRL) in the evaluated environment and provide a foundation for more complex future applications of spiking networks. [ABSTRACT FROM AUTHOR]

Cluster analysis is one important field in pattern recognition and machine learning, consisting in an attempt to distribute a set of data patterns into groups, considering only the inner properties of those data. One of the most popular techniques for data clustering is the K-Means algorithm, due to its simplicity and easy implementation. But K-Means is strongly dependent on the initial point of the search, what may lead to suboptima (local optima) solutions. In the past few decades, Evolutionary Algorithms (EAs), like Group Search Optimization (GSO), have been adapted to the context of cluster analysis, given their global search capabilities and flexibility to deal with hard optimization problems. However, given their stochastic nature, EAs may be slower to converge in comparison to traditional clustering models (like K-Means). In this work, three hybrid memetic approaches between K-Means and GSO are presented, named FMKGSO, MKGSO and TMKGSO, in such a way that the global search capabilities of GSO are combined with the fast local search performances of K-Means. The degree of influence of K-Means on the behavior of GSO method is evaluated by a set of experiments considering both real-world problems and synthetic data sets, using five clustering metrics to access how good and robust the proposed hybrid memetic models are. [ABSTRACT FROM AUTHOR]

In the field of time series forecasting, the combination of linear and nonlinear models has been explored to improve the accuracy of predictions when compared with the performance of individual models. Traditional forecasting methods such as the autoregressive integrated moving average (ARIMA) and exponential smoothing (ETS) are employed to map linear patterns in a time series while neural networks (ANNs) and support vector machines (SVMs) map nonlinear patterns. A traditional hybrid ARIMA–ANN system works by performing linear relationships in the data, considering the residual error a nonlinear component which is mapped by the ANN. The nature of a time series can be taken into consideration before applying a model. This work employs both linear and nonlinear patterns of a time series based on its volatility. This is explored using a hybrid evolutionary system composed by a simple exponential smoothing filter, ARIMA and autoregressive (AR) linear models and a SVR model. Particle swarm optimization is employed to optimize the order of the AR model, SVR parameters and number of lags of the time series. Experimental results show that the evolutionary hybrid system presented promising results in the forecasting domain. [ABSTRACT FROM AUTHOR]

The theory of quantum algorithms promises unprecedented benefits of harnessing the laws of quantum mechanics for solving certain computational problems. A prerequisite for applying quantum algorithms to a wide range of real-world problems is loading classical data to a quantum state. Several circuit-based methods have been proposed for encoding classical data as probability amplitudes of a quantum state. However, in these methods, either quantum circuit depth or width must grow linearly with the data size, nullifying the advantage of representing exponentially many classical data in a quantum state. In this paper, we present a configurable bidirectional procedure that addresses this problem by tailoring the resource trade-off between quantum circuit width and depth. In particular, we show a configuration that encodes an N-dimensional classical data using a quantum circuit whose width and depth both grow sublinearly with N. We demonstrate proof-of-principle implementations on five quantum computers accessed through the IBM and IonQ quantum cloud services. [ABSTRACT FROM AUTHOR]

Abstract: In recent years, the Extreme Learning Machine (ELM) has been hybridized with the Particle Swarm Optimization (PSO) and such hybridization is called PSO-ELM. In most of these hybridizations, the PSO uses the Global topology. However, other topologies were designed to improve the performance of the PSO. In the literature, it is well known that the performance of the PSO depends on its topology, and there is not a best topology for all problems. Thus, in this paper, we investigate the effect of eight PSO topologies on performance of the PSO-ELM. The results showed empirically that the Global topology was more promising than all other topologies in optimizing the PSO-ELM according to the root mean squared error (RMSE) on the validation set in most of the evaluated datasets. However, no correlation was detected between this good performance on the RMSE and the testing accuracy. [Copyright &y& Elsevier]

We present here a work that applies an automatic construction of ensembles based on the Clustering and Selection (CS) algorithm for time series forecasting. The automatic method, called CSELM, initially finds an optimum number of clusters for training data set and subsequently designates an Extreme Learning Machine (ELM) for each cluster found. For model evaluation, the testing data set are submitted to clustering technique and the nearest cluster to data input will give a supervised response through its associated ELM. Self-organizing maps were used in the clustering phase. Adaptive differential evolution was used to optimize the parameters and performance of the different techniques used in the clustering and prediction phases. The results obtained with the CSELM method are compared with results obtained by other methods in the literature. Five well-known time series were used to validate CSELM. [ABSTRACT FROM AUTHOR]

Highlights: [•] We present a method for optimization of the activation functions for ANN. [•] The proposed optimization method uses Simulated Annealing and Tabu Search. [•] The proposed method is good for forecasting distinct time series. [ABSTRACT FROM AUTHOR]

Abstract: Social class differences in the prevalence of Common Mental Disorder (CMD) are likely to vary according to time, culture and stage of economic development. The present study aimed to investigate the use of optimization of architecture and weights of Artificial Neural Network (ANN) for identification of the factors related to CMDs. The identification of the factors was possible by optimizing the architecture and weights of the network. The optimization of architecture and weights of ANNs is based on Particle Swarm Optimization with early stopping criteria. This approach achieved a good generalization control, as well as similar or better results than other techniques, but with a lower computational cost, with the ability to generate small networks and with the advantage of the automated architecture selection, which simplify the training process. This paper presents the results obtained in the experiments with ANNs in which it was observed an average percentage of correct classification of individuals with positive diagnostic for the CMDs of 90.59%. [Copyright &y& Elsevier]

Abstract: Reservoir computing is a framework for computation like a recurrent neural network that allows for the black box modeling of dynamical systems. In contrast to other recurrent neural network approaches, reservoir computing does not train the input and internal weights of the network, only the readout is trained. However it is necessary to adjust parameters to create a “good” reservoir for a given application. In this study we introduce a method, called RCDESIGN (reservoir computing and design training). RCDESIGN combines an evolutionary algorithm with reservoir computing and simultaneously looks for the best values of parameters, topology and weight matrices without rescaling the reservoir matrix by the spectral radius. The idea of adjust the spectral radius within the unit circle in the complex plane comes from the linear system theory. However, this argument does not necessarily apply to nonlinear systems, which is the case of reservoir computing. The results obtained with the proposed method are compared with results obtained by a genetic algorithm search for global parameters generation of reservoir computing. Four time series were used to validate RCDESIGN. [Copyright &y& Elsevier]

A time series is a sequence of observations of a random variable. Hence, it is a stochastic process. Forecasting time series data is important component of operations research because these data often provide the foundation for decision models. This models are used to predict data points before they are measured based on known past events. Researches in this subject have been done in many areas like economy, energy production, ecology and others. To improve the process of time series forecasting it is important to identify which of past values will be considered to be used in the models by eliminating redundant or irrelevant attributes. Two hybrid systems Harmony Search with Neural Networks (HS) and Temporal Memory Search with Neural Networks (TMS) are improved and a new one is proposed: the Temporal Memory Search Limited with Neural Networks (TMSL). The performance of the techniques is investigated through an empirical evaluation on twenty real-world time series. [ABSTRACT FROM AUTHOR]

The performance of an artificial neural network (ANN) depends upon the selection of proper connection weights, network architecture, and cost function during network training. This paper presents a hybrid approach (GaTSa) to optimize the performance of the ANN in terms of architecture and weights. GaTSa is an extension of a previous method (TSa) proposed by the authors. GaTSa is based on the integration of the heuristic simulated annealing (SA), tabu search (TS), genetic algorithms (GA), and backpropagation, whereas TSa does not use GA. The main advantages of GaTSa are the following: a constructive process to add new nodes in the architecture based on GA, the ability to escape from local minima with uphill moves (SA feature), and faster convergence by the evaluation of a set of solutions (TS feature). The performance of GaTSa is investigated through an empirical evaluation of 11 public-domain data sets using different cost functions in the simultaneous optimization of the multilayer perceptron ANN architecture and weights. Experiments demonstrated that GaTSa can also be used for relevant feature selection. GaTSa presented statistically relevant results in comparison with other global and local optimization techniques. [ABSTRACT FROM PUBLISHER]

Abstract: The use of artificial neural networks implies considerable time spent choosing a set of parameters that contribute toward improving the final performance. Initial weights, the amount of hidden nodes and layers, training algorithm rates and transfer functions are normally selected through a manual process of trial-and-error that often fails to find the best possible set of neural network parameters for a specific problem. This paper proposes an automatic search methodology for the optimization of the parameters and performance of neural networks relying on use of Evolution Strategies, Particle Swarm Optimization and concepts from Genetic Algorithms corresponding to the hybrid and global search module. There is also a module that refers to local searches, including the well-known Multilayer Perceptrons, Back-propagation and the Levenberg–Marquardt training algorithms. The methodology proposed here performs the search using the aforementioned parameters in an attempt to optimize the networks and performance. Experiments were performed and the results proved the proposed method to be better than trial-and-error and other methods found in the literature. [Copyright &y& Elsevier]

Abstract: This paper introduces an approach called Clustering and Co-evolution to Construct Neural Network Ensembles (CONE). This approach creates neural network ensembles in an innovative way, by explicitly partitioning the input space through a clustering method. The clustering method allows a reduction in the number of nodes of the neural networks that compose the ensemble, thus reducing the execution time of the learning process. This is an important characteristic especially when evolutionary algorithms are used. The clustering method also ensures that different neural networks specialize in different regions of the input space, working in a divide-and-conquer way, to maintain and improve the accuracy. Besides, the clustering method facilitates the understanding of the system and makes a straightforward distributed implementation possible. The experiments performed with seven classification databases and three different co-evolutionary algorithms show that CONE considerably reduces the execution time without prejudicing (and even improving) the accuracy, even when a distributed implementation is not used. [Copyright &y& Elsevier]

Meta-Learning has been successfully applied to acquire knowledge used to support the selection of learning algorithms. Each training example in Meta-Learning (i.e. each meta-example) is related to a learning problem and stores the experience obtained in the empirical evaluation of a set of candidate algorithms when applied to the problem. The generation of a good set of meta-examples can be a costly process depending for instance on the number of available learning problems and the complexity of the candidate algorithms. In this work, we proposed the Active Meta-Learning, in which Active Learning techniques are used to reduce the set of meta-examples by selecting only the most relevant problems for meta-example generation. In an implemented prototype, we evaluated the use of two different Active Learning techniques applied in two different Meta-Learning tasks. The performed experiments revealed a significant gain in the Meta-Learning performance when the active techniques were used to support the meta-example generation. [ABSTRACT FROM AUTHOR]

This paper introduces a methodology for neural net- work global optimization. The aim is the simultaneous optimization of multilayer perceptron (MLP) network weights and architectures, in order to generate topologies with few connections and high classification performance for any data sets. The approach combines the advantages of simulated annealing, tabu search and the backpropagation training algorithm in order to generate an automatic process for producing networks with high classification performance and low complexity. Experimental results obtained with four classification problems and one prediction problem has shown to be better than those obtained by the most commonly used optimization techniques. [ABSTRACT FROM AUTHOR]

In this letter, the computational power of a class of random access memory (RAM)-based neural networks, called general single-layer sequential weightless neural networks (GSSWNNs), is analyzed. The theoretical results presented, besides helping the understanding of the temporal behavior of these networks, could also provide useful insights for the developing of new learning algorithms. [ABSTRACT FROM AUTHOR]

This work examines the use of Hybrid Intelligent Systems in the pattern recognition system of an artificial nose. The connectionist approaches Multi-Layer Perceptron and Time Delay Neural Networks, and the hybrid approaches Feature-Weighted Detector and Evolving Neural Fuzzy Networks were investigated. A Wavelet Filter is evaluated as a preprocessing method for odor signals. The signals generated by an artificial nose were composed by an array of conducting polymer sensors and exposed to two different odor databases. [ABSTRACT FROM AUTHOR]

We present here an original work that applies meta-learning approaches to select models for time-series forecasting. In our work, we investigated two meta-learning approaches, each one used in a different case study. Initially, we used a single machine learning algorithm to select among two models to forecast stationary time series (case study I). Following, we used the NOEMON approach, a more recent work in the meta-learning area, to rank three models used to forecast time series of the M3-Competition (case study II). The experiments performed in both case studies revealed encouraging results. [Copyright &y& Elsevier]

The selection of a good model for forecasting a time series is a task that involves experience and knowledge. Employing machine learning algorithms is a promising approach to acquiring knowledge in regards to this task. A supervised classification method originating from the symbolic data analysis field is proposed for the model selection problem. This method was applied in the task of selecting between two widespread models, and compared to other learning algorithms. To date, it has obtained the lowest classification errors among all the tested algorithms. [Copyright &y& Elsevier]

This paper presents an approach of using Simulated Annealing and Tabu Search for the simultaneous optimization of neural network architectures and weights. The problem considered is the odor recognition in an artificial nose. Both methods have produced networks with high classification performance and low complexity. Generalization has been improved by using the backpropagation algorithm for fine tuning. The combination of simple and traditional search methods has shown to be very suitable for generating compact and efficient networks. [ABSTRACT FROM AUTHOR]

Neuronal groups projecting widely in the brain are being experimentally associated to attention and mood changes. Those groups are known to exert a modulatory effect over other larger groups. On the other hand, some people think of the brain functions as being performed by specialized modular systems. In this work, we propose an architecture of modular nature to explore a particular decision process. We show the importance of the modulatory effect of a special evaluation segment in that process. [ABSTRACT FROM AUTHOR]

Cites several abstracts on neural systems submitted at the VIIth Brazilian Symposium on Artificial Neural Networks held in 2002 at Porto De Galinhas-Pernambuco. Ways on how to deal with non-orthogonal signals by Allan Kardex Barros, Andrzej Cichocki and Noboru Ohnishi; Neuro-symbolic language for monotonic and non-monotonic parallel logical inference by means of artificial neural networks; Performance of fuzzy combination schemes.

Magnetic skyrmions are nanoscale windings of the spin structure that can be observed in chiral magnets and hold promise for potential applications in storing or processing information. Pinning due to ever-present material imperfections crucially affects the mobility of skyrmions. Therefore, a proper understanding of how magnetic skyrmions pin to defects is necessary for the development and performance of spintronic devices. Here we present a fundamental analysis on the interactions of single skyrmions with atomic defects of distinctly different origins, in a Co monolayer on Pt, based on minimum-energy paths considerations and atomic-spin simulations. We first report the preferred pinning loci of the skyrmion as a function of its nominal size and the type of defect being considered, to further reveal the manipulation and "breathing" of skyrmion core in the vicinity of a defect. We also show the behavior of skyrmions in the presence of an extended defect of particular geometry, that can lead to ratcheted skyrmion motion or a facilitated guidance on a defect "trail." We close the study with reflections on the expected thermal stability of the skyrmion against collapse on itself for a given nature of the defect, and discuss the applications where control of skyrmions by defects is of particular interest. [ABSTRACT FROM AUTHOR]

A large number of classification algorithms have been proposed in the machine learning literature. These algorithms have different pros and cons, and no algorithm is the best for all datasets. Hence, a challenging problem consists of choosing the best classification algorithm with its best hyper-parameter settings for a given input dataset. In the last few years, Automated Machine Learning (Auto-ML) has emerged as a promising approach for tackling this problem, by doing a heuristic search in a large space of candidate classification algorithms and their hyper-parameter settings. In this work we propose an improved version of our previous Evolutionary Algorithm (EA) – more precisely, an Estimation of Distribution Algorithm – for the Auto-ML task of automatically selecting the best classifier ensemble and its best hyper-parameter settings for an input dataset. The new version of this EA was compared against its previous version, as well as against a random forest algorithm (a strong ensemble algorithm) and a version of the well-known Auto-ML method Auto-WEKA adapted to search in the same space of classifier ensembles as the proposed EA. In general, in experiments with 21 datasets, the new EA version obtained the best results among all methods in terms of four popular predictive accuracy measures: error rate, precision, recall and F-measure. [ABSTRACT FROM AUTHOR]

Digital assets termed cryptocurrencies are creating new paradigms for financial transactions as well as alternative means of capital. Despite their increasing relevance in financial situations, a complete understanding of the entire system remains lacking. Here the cryptocurrency market is treated as a complex system and analyzed using methods from statistical physics. The complexity–entropy causality plane (or CH plane) is employed in order to explore disorder and complexity in the space of cryptocurrencies. They are found to exist on distinct planar locations in the representation space, ranging from structured to stochastic-like behavior. The temporal trajectories of entropy and statistical complexity of prices vary drastically with position along the plane. Lastly cryptocurrencies appear to be characterized by ordinal patterns that represent strictly decreasing or increasing trends in price. The present analysis expands the understanding of and helps to quantify varying degrees of complexity in cryptocurrencies. • We map cryptocurrencies to the complexity–entropy causality plane. • We discover rather distinct degrees of complexity in the cryptocurrency space. • Their price behavior ranges from deterministic chaos to stochastic-like processes. • Most cryptocurrencies are characterized by downward or upward trends. [ABSTRACT FROM AUTHOR]

Abstract The cryptocurrency market has emerged in recent years with profound implications to central financial institutions such as banking. Bitcoin and other popular cryptocurrencies were recently analyzed using methods from statistical physics, but focus was exclusively on behavior of prices. Here we explore multifractal behavior of daily price changes and daily volume changes of fifty cryptocurrencies using the multifractal detrended fluctuation analysis (MF-DFA). Our results indicate that price changes are more complex than volume changes, and that large and small fluctuations dominate multifractal behavior of price and volume changes, respectively. We also find there is an absence of correlations in price changes, while volume changes present anti-persistent long-term correlations. Shuffling the series reveals that multifractality of both price and volume changes arises from a broad probability density function. The multifractal behavior of the cryptocurrency market is surprisingly similar to that of stock markets, but differs from that of regular exchange rates. Highlights • We discover common multifractal behavior in cryptocurrency exchanges. • Price changes and volume changes follow different multifractal dynamics. • Cryptocurrency market is similar to stock market but differs from regular currencies. [ABSTRACT FROM AUTHOR]

Digital assets termed cryptocurrencies are correlated. We analyze cross correlations between price changes of different cryptocurrencies using methods of random matrix theory and minimum spanning trees. We find that the cross correlation matrix exhibits non-trivial hierarchical structures and groupings of cryptocurrency pairs, which are not present in the partial cross correlations. In sharp contrast to the predictions for other financial markets, we discover that most of the eigenvalues in the spectrum of the cross correlation matrix do not agree with the universal predictions of random matrix theory, but the few of the largest eigenvalues deviate as expected. The minimum spanning tree of cryptocurrency cross correlations reveals distinct community structures that are surprisingly stable. Collective behaviors that are present in the cryptocurrency market can be useful for the construction of portfolio of cryptocurrencies as well as for future research on the subject. [ABSTRACT FROM AUTHOR]

Cryptocurrencies represent a new type of financial assets that are traded in a decentralized and transparent way. Recently, cryptocurrencies with large market capitalization (mostly Bitcoin) have been studied theoretically, but a deeper understanding of their underlying mechanisms remains elusive. Here we explore the nonextensivity of price changes for 20 cryptocurrency exchanges from 2013 until 2017. We discover nonextensive triplets in the cryptocurrency market, where the three associated values for Bitcoin are remarkably close to those of the logistic map near the edge of chaos. The current findings strongly indicate that the cryptocurrency market represents a system whose physics is properly described by nonextensive statistical mechanics. Our results shed light on the complex and volatile nature of cryptocurrencies, and establish the first formal link with the nonextensive theory. [ABSTRACT FROM AUTHOR]

An introduction is presented in which the guest editors discuss various reports within the issue on topics including selection of algorithms to solve traveling salesman problem using meta-learning, selecting features to improve time series forecasting using hybrid intelligent techniques, and comparison of single classification methods and classifier ensembles performances.

Presents information on the papers selected from the VIIth Brazilian Symposium on Neural Networks proceedings, held in 2002. Presentation of a novel class of recurrent neural fuzzy networks; Proposed constructive and pruning methods; Investigation of artificial neural networks.

Researches in natural neuron dynamics have shown that phase transition and chaos provide optimal behaviour for information processing. In artificial neural models that behavior is expected also to maximize the neuron learnability. By employing an open quantum systems approach, we propose a new quantum information extraction method in the quantum RAM node dynamics where complex values are iterated. We experimentally show bifurcation and chaos emergence by varying their parameters. [ABSTRACT FROM AUTHOR]

Time series forecasting is an important research field in machine learning. Since the literature shows several techniques for the solution of this problem, combining outputs of different models is a simple and robust strategy. However, even when using combiners, the experimenter may face the following dilemma: which technique should one use to combine the individual predictors? Inspired by classification and pattern recognition algorithms, this work presents a dynamic selection method of forecast combiners. In the dynamic selection, each test pattern is submitted to a certain combiner according to a nearest neighbor rule. The proposed method was used to forecast eight time series with chaotic behavior in short and long term. In general, the dynamic selection presented satisfactory results for all datasets. [ABSTRACT FROM AUTHOR]

Algorithm selection is an important task in different domains of knowledge. Meta-learning treats this task by adopting a supervised learning strategy. Training examples in meta-learning (called meta-examples) are generated from experiments performed with a pool of candidate algorithms in a number of problems, usually collected from data repositories or synthetically generated. A meta-learner is then applied to acquire knowledge relating features of the problems and the best algorithms in terms of performance. In this paper, we address an important aspect in meta-learning which is to produce a significant number of relevant meta-examples. Generating a high quality set of meta-examples can be difficult due to the low availability of real datasets in some domains and the high computational cost of labelling the meta-examples. In the current work, we focus on the generation of meta-examples for meta-learning by combining: (1) a promising approach to generate new datasets (called datasetoids ) by manipulating existing ones; and (2) active learning methods to select the most relevant datasets previously generated. The datasetoids approach is adopted to augment the number of useful problem instances for meta-example construction. However not all generated problems are equally relevant. Active meta-learning then arises to select only the most informative instances to be labelled. Experiments were performed in different scenarios, algorithms for meta-learning and strategies to select datasets. Our experiments revealed that it is possible to reduce the computational cost of generating meta-examples, while maintaining a good meta-learning performance. [ABSTRACT FROM AUTHOR]

Training artificial neural networks requires a tedious empirical evaluation to determine a suitable neural network architecture. To avoid this empirical process several techniques have been proposed to automatise the architecture selection process. In this paper, we propose a method to perform parameter and architecture selection for a quantum weightless neural network (qWNN). The architecture selection is performed through the learning procedure of a qWNN with a learning algorithm that uses the principle of quantum superposition and a non-linear quantum operator. The main advantage of the proposed method is that it performs a global search in the space of qWNN architecture and parameters rather than a local search. [ABSTRACT FROM AUTHOR]

Abstract: A supervised learning algorithm for quantum neural networks (QNN) based on a novel quantum neuron node implemented as a very simple quantum circuit is proposed and investigated. In contrast to the QNN published in the literature, the proposed model can perform both quantum learning and simulate the classical models. This is partly due to the neural model used elsewhere which has weights and non-linear activations functions. Here a quantum weightless neural network model is proposed as a quantisation of the classical weightless neural networks (WNN). The theoretical and practical results on WNN can be inherited by these quantum weightless neural networks (qWNN). In the quantum learning algorithm proposed here patterns of the training set are presented concurrently in superposition. This superposition-based learning algorithm (SLA) has computational cost polynomial on the number of patterns in the training set. [Copyright &y& Elsevier]

Interest in hybrid methods that combine artificial neural networks and fuzzy inference systems has grown in recent years. These systems are robust solutions that search for representations of domain knowledge, reasoning on uncertainty, automatic learning and adaptation. However, the design and definition of the parameter effectiveness of such systems is still a hard task. In the present work, we perform a statistical analysis to verify interactions and interrelations between parameters in the design of neuro-fuzzy systems. The analysis is carried out using a powerful statistical tool, namely, Design of Experiments (DOE), in two neuro-fuzzy models — Adaptive Neuro Fuzzy Inference System (ANFIS) and Evolving Fuzzy Neural Networks (EFuNN). The results show that, for ANFIS, input MFs number and output MFs shape are usually the factors with the largest influence on the system's RMSE. For EFFuNN, the MF shape and the interaction between MF shape and number usually have the largest effect size. [ABSTRACT FROM AUTHOR]

Abstract: This paper presents approaches to interval-valued time series forecasting. The first and second approaches are based on the autoregressive (AR) and autoregressive integrated moving average (ARIMA) models, respectively. The third approach is based on an artificial neural network (ANN) model and the last is based on a hybrid methodology that combines both ARIMA and ANN models. Each approach fits, respectively, two models on the mid-point and range of the interval values assumed by the interval-valued time series in the learning set. The forecasting of the lower and upper bounds of the interval value of the time series is accomplished through a combination of forecasts from the mid-point and range of the interval values. The evaluation of the models presented is based on the estimation of the average behavior of the mean absolute error and mean squared error in the framework of a Monte Carlo experiment. The results demonstrate that the approaches are useful in forecasting alternatives for interval-valued time series and indicate that the hybrid model is an effective way to improve the forecasting accuracy achieved by any one of the models separately. [Copyright &y& Elsevier]

A novel way to simulate Turing Machines (TMs) by Artificial Neural Networks (ANNs) is proposed. We claim that the proposed simulation is in agreement with the correct interpretation of Turing's analysis of computation; compatible with the current approaches to analyze cognition as an interactive agent-environment process; and physically realizable since it does not use connection weights with unbounded precision. A full description of an implementation of a universal TM into a recurrent sigmoid ANN focusing on the TM finite state control is given, leaving the tape, an infinite resource, as an external non-intrinsic feature. Also, motivated by the results on the limit of what can actually be computed by ANNs when noise is taken into account, we introduce the notion of Definite Turing Machine and investigate some of its properties. [ABSTRACT FROM AUTHOR]

This work addresses the problem of finding the adjustable parameters of a learning algorithm using Genetic Algorithms. This problem is also known as the model selection problem. In this paper, some model selection techniques (e.g., crossvalidation and bootstrap) are used as objective functions of a Genetic Algorithm. The Genetic Algorithm is modified in order to allow the efficient use of these objective functions by means of occam's razor, growing, and other heuristics. Some modifications explore intrinsic features of Genetic Algorithms, such as their ability to handle multiple and noise objective functions. The proposed techniques are very general and may be applied to a large range of learning algorithms. [ABSTRACT FROM AUTHOR]

A fundamental question in the field of artificial neural networks is what set of problems a given class of networks can perform (computability). Such a problem can be made less general, but no less important, by asking what these networks could learn by using a given training procedure (learnability). The basic purpose of this paper is to address the learnability problem. Specifically, it analyses the learnability of sequential RAM-based neural networks. The analytical tools used are those of Automata Theory. In this context, this paper establishes which class of problems and under what conditions such networks, together with their existing learning rules, can learn and generalise. This analysis also yields techniques for both extracting knowledge from and inserting knowledge into the networks. The results presented here, besides helping in a better understanding of the temporal behaviour of sequential RAM-based networks, could also provide useful insights for the integration of the symbolic/connectionist paradigms. [ABSTRACT FROM AUTHOR]

We present a quantum probabilistic associative memory using the inverse of quantum Fourier transform and Grover's algorithm to recover existing or similar patterns in the memory. The content of the memory is created using a generator of a superposition state representing a given set of patterns. We discuss the architecture of the proposed memory including the storing, recovering and processing of similarity tolerance of the input query. The associative memory can extrapolate and recover similar stored patterns. The system is unitary and runs in O (n) steps, where n is the number of qubits of the patterns. [ABSTRACT FROM AUTHOR]

Millions of people throughout the world describe themselves as being deaf. Some of them suffer from severe hearing loss and consequently use an alternative manner with which to communicate with society by means of either written or visual language. There are several sign languages capable of dealing with such a need. Nonetheless, a communication gap still exists even when using such languages, since only a small fraction of the population is able to use them. Over the last few years, due to the increasing need for universal accessibility when using computational resources, gesture recognition has been widely researched. Thus, in an attempt to reduce this communication gap, our approach proposes a computational solution in order to translate static gesture symbols into text symbols, through computer vision, without the use of hand sensors or gloves. In order to guarantee the highest quality, with emphasis on the reliability of the system and real-time translation, we have developed an approach based on the Extreme Learning Machine (ELM) pattern recognition algorithms fully implemented in hardware, and have assessed it to measure these two metrics. Hardware components were designed in order to perform the best image processing and pattern recognition tasks used within the project. As a case study, and so as to validate the technique, a recognition system for the Brazilian Sign Language (LIBRAS) was implemented. Besides ensuring that this approach could be used for any static hand gesture symbol recognition, our main goal was to guarantee fast, reliable gesture recognition for communication between humans. Experimental results have demonstrated that the system is able to recognize LIBRAS symbols with an accuracy of 97%, a response time of 6.5ms per letter recognition, and using only 43% (about 64,851 logic elements) of the FPGA area. [ABSTRACT FROM AUTHOR]