Nature-Inspired Design of Hybrid Intelligent Systems

In this paper, we present the advantage of using a general type-2 fuzzy edge detector method in the preprocessing phase of a face recognition system. The Sobel and Prewitt edge detectors combined with GT2 FSs are considered in this work. In our approach, the main idea is to apply a general type-2 fuzzy edge detector on two image databases to reduce the size of the dataset to be processed in a face recognition system. The recognition rate is compared using different edge detectors including the fuzzy edge detectors (type-1 and interval type-2 FS) and the traditional Prewitt and Sobel operators.

As Granular Computing has gained interest, more research has lead into using different representations for Information Granules, i.e., rough sets, intervals, quotient space, fuzzy sets; where each representation offers different approaches to information granulation. These different representations have given more flexibility to what information granulation can achieve. In this overview paper, the focus is only on journal papers where Granular Computing is studied when fuzzy logic systems are used, covering research done with Type-1 Fuzzy Logic Systems, Interval Type-2 Fuzzy Logic Systems, as well as the usage of general concepts of Fuzzy Systems.

In this paper, two bio-inspired methods are applied to optimize the type-2 and type-1 fuzzy integrator used in the neural network with fuzzy weights. The genetic algorithm and particle swarm optimization are used to optimize the type-2 and type-1 fuzzy integrator that work in the integration of the output for the ensemble neural network with three networks. One neural network uses type-2 fuzzy inference systems with Gaussian membership functions to obtain the fuzzy weights; the second neural network uses type-2 fuzzy inference systems with triangular membership functions; and the third neural network uses type-2 fuzzy inference systems with triangular membership functions with uncertainty in the standard deviation. In this work, an optimized type-2 and type-1 fuzzy integrator to manage the output of the ensemble neural network and the results for the two bio-inspired methods are presented. The proposed approach is applied to a case of time series prediction, specifically in Mackey-Glass time series.

In this paper, we present optimization of the Interval Type-2 Fuzzy Possibilistic C-Means (IT2FPCM) algorithm using Particle Swarm Optimization (PSO), with the goal of automatically finding the optimal number of clusters and the optimal lower and upper limit of Fuzzy and Possibility exponents of weight of the of the IT2FPCM algorithm, and also the centroids of clusters of each dataset tested with the IT2FPCM algorithm optimized using PSO.

In this paper, a method for edge detection in digital images based on morphological gradient technique in combination with Choquet integral, and the interval type-2 Choquet integral is proposed. The aggregation operator is used as a method to integrate the four gradients of the edge detector. Simulation results with real images and synthetic images are presented and the results show that the interval type-2 Choquet integral is able to improve the detected edge.

Google AdWords has a bidding price optimization method for its campaigns, where the user establishes the maximum bidding price, and AdWords adapts the final bidding price according to the performance of a campaign. This chapter proposes a bidding price controller based on a fuzzy inference system. Specifically, two approaches are considered: a type-1 fuzzy inference system, and an interval type-2 fuzzy inference system. The results show that the proposed methods are superior to the AdWords optimization method, and that there is not enough statistical evidence to support the superiority of the interval type-2 fuzzy inference system against the type-1 fuzzy inference system, although type-2 is slightly better.

This work proposes an approach for graphically representing intuitionistic fuzzy sets for their use in Mamdani fuzzy inference systems. The proposed approach is used and plots for several membership and non-membership functions are presented, including: triangular, Gaussian, trapezoidal, generalized bell, sigmoidal, and left-right functions. Plots of some operators used in fuzzy logic are also presented, i.e., union, intersection, implication, and alpha-cut operators. The proposed approach should produce plots that are clear to understand in the design of an intuitionistic fuzzy inference system, as the membership and non-membership functions are clearly separated and can be plotted in the same figure and still be recognized with ease.

In this paper, we are presenting a modification of the Gravitational Search Algorithm (GSA) using type-2 fuzzy logic to dynamically change the alpha parameter and provide a different gravitation and acceleration to each agent in order to improve its performance. We test this approach with benchmark mathematical functions. Simulation results show the advantage of the proposed approach.

This paper describes the construction of intelligent hybrid architectures and the optimization of the fuzzy integrators for time series prediction; interval type-2 fuzzy neural networks (IT2FNN). IT2FNN used hybrid learning algorithm techniques (gradient descent backpropagation and gradient descent with adaptive learning rate backpropagation). The IT2FNN is represented by Takagi–Sugeno–Kang reasoning. Therefore this TSK IT2FNN is represented as an adaptive neural network with hybrid learning in order to automatically generate an interval type-2 fuzzy logic system (TSK IT2FLS). We use interval type-2 and type-1 fuzzy systems to integrate the output (forecast) of each Ensemble of ANFIS models. Particle Swarm Optimization (PSO) was used for the optimization of membership functions (MFs) parameters of the fuzzy integrators. The Mackey-Glass time series is used to test of performance of the proposed architecture. Simulation results show the effectiveness of the proposed approach.

In the past years, efforts have been made to improve the efficiency of long-term time series forecasting. However, when the involved series is highly oscillatory and nonlinear, this is still an open problem. Given the fact that signals may be approximated as linear combinations of sine functions, the study of the behavior of an adaptive dynamical model able to reproduce a sine function may be relevant for long-term prediction. In this chapter, we present an analysis of the modeling and prediction abilities of the “Long Short-Term Memory” (LSTM) recurrent neural network, when the input signal has a discrete sine function shape. Previous works have shown that LSTM is able to learn relevant events among long-term lags, however, its oscillatory abilities have not been analyzed enough. In our experiments, we found that some configurations of LSTM were able to model the signal, accurately predicting up to 400 steps forward. However, we also found that similar architectures did not perform properly when experiments were repeated, probably due to the fact that the LSTM architectures got over trained and the learning algorithm got trapped in a local minimum.

This chapter presents a pulse-coupled neural network architecture, PCNN, to segment imagery acquired with UAV images. The images correspond to normalized difference vegetation index values. The chapter describes the image analysis system design, the image acquisition elements, the original PCNN architect, the simplified PCNN, the automatic parameter setting methodology, and qualitative and quantitative results of the proposed method using real aerial images.

In this paper, a new model for arrhythmia classification using a modular LVQ neural network architecture and a type-2 fuzzy system is presented. This work focuses on the implementation of a type-2 fuzzy system to determine the shortest distance in a LVQ neural network competitive layer. In this work, the MIT-BIH arrhythmia database with 15 classes was used. Results show that using five modules architecture could be a good approach for classification of arrhythmias.

In this paper, a method is proposed to diagnose the blood pressure of a patient (Astolic pressure, diastolic pressure, and pulse). This method consists of a modular neural network and its response with average integration. The proposed approach consists on applying these methods to find the best architecture of the modular neural network and the lowest prediction error. Simulations results show that the modular network produces a good diagnostic of the blood pressure of a patient.

Free vibration occurs when a mechanical system is disturbed from equilibrium by an external force and then it is allowed to vibrate freely. In free vibrations, the system oscillates under the influence of inherent forces on the system itself. Free vibrations are associated with natural frequencies that are properties of the oscillating system, quantified in parameters such as mass, shape, and stiffness distribution. A number of these mechanical characteristics can be inferred from vibration patterns or from the generated sound using the adequate sensors. It is well known that liquid level inside a container modifies its natural frequencies. Unfortunately, other container characteristics such as shape, composition, temperature, and pressure modifies the natural frequencies of vibration making the task of level measurement nontrivial. Preliminary experiments aiming to do measurement of liquid content level and container characterization are presented in this work. Spectral analysis in Fourier domain is used to perform feature extraction, with the feature vectors containing information about the frequencies having the greatest amplitude in the respective spectral analysis. Classification has been carried out using two computational intelligence techniques for comparison purposes: neural network classification and a fuzzy logic inference system built using singleton fuzzifier, product inference rule, Gaussian membership functions and center average defuzzifier. Preliminary results showed a better performance when using the neural network-based approach in comparison to the fuzzy logic-based approach, obtaining in average a MSE of 0.02 and 0.09, respectively.

In this paper a comparison of optimization techniques for a Modular Neural Network (MNN) with a granular approach is presented. A Hierarchical Genetic Algorithm, a Firefly Algorithm (FA), and a Grey Wolf Optimizer are developed to perform a comparison of results. These algorithms design optimal MNN architectures, where their main task is the optimization of some parameters of MNN such as, number of sub modules, percentage of information for the training phase and number of hidden layers (with their respective number of neurons) for each sub module and learning algorithm. The MNNs are applied to human recognition based on iris biometrics, where a benchmark database is used to perform the comparison, having as objective function in each optimization algorithm the minimization of the error of recognition.

In this paper, a modular neural network (MNN) architecture based on competitive clustering and a winner-takes-all strategy is proposed. In this case, the modules are obtained from clustering the training data with a competitive layer. And each module consists of a single hidden layer nonlinear autoregressive neural network. This MNN architecture can be used for short-term and long-term time series forecasting.

This paper proposes an improvement to the algorithm differential evolution (DE) using fuzzy logic. The main contribution of this work is to dynamically adapt the parameter of mutation (F) using a fuzzy system, with the aim that the fuzzy system calculates the optimal parameters of the DE algorithm for obtaining better solutions, in this way arriving to the proposed new fuzzy differential evolution (FDE) algorithm. In this paper, experiments are performed with a set of mathematical functions using the proposed method to show the advantages of the FDE algorithm.

This paper develops a new fuzzy harmony search algorithm (FHS) for solving optimization problems. FHS employs a novel method using fuzzy logic for adaptation of parameter the pitch adjustment (PArate) that enhances accuracy and convergence of harmony search (HS) algorithm. In this paper the impact of constant parameters on harmony search algorithm is discussed and a strategy for tuning these parameters is presented. The FHS algorithm has been successfully applied to various benchmarking optimization problems. Numerical results reveal that the proposed algorithm can find better solutions when compared to HS and other heuristic methods and is a powerful search algorithm for various benchmarking optimization problems.

The firefly algorithm is a bioinspired metaheuristic-based on the firefly’s behavior. This paper shows previous works on parameters analysis and dynamical parameter adjustment, using different approaches and fuzzy logic.

This paper describes the enhancement of the water cycle algorithm (WCA) using a fuzzy inference system to dynamically adapt its parameters. The original WCA is compared in terms of performance with the proposed method called WCA with dynamic parameter adaptation (WCA-DPA). Simulation results on a set of well-known test functions show that the WCA is improved with a fuzzy dynamic adaptation of the parameters.

The main goal of this paper is to improve the performance of the fireworks algorithm (FWA). This improvement is based on fuzzy logic, which means we implemented different fuzzy inference systems into the FWA with the intent to convert parameters that were usually constant in dynamic parameters. After having studied the performance of the FWA, we concluded that two parameters are key of the performance the algorithm (FWA), the parameters that we comment are: the number of sparks and explosion amplitude of each firework, these parameters were adjusted using fuzzy logic, and this adjustment we called Fuzzy Fireworks Algorithm and we denoted as FzFWA. We can justify this adjustment of parameters with simulation results obtained in evaluating six mathematical benchmark functions.

In this paper, we describe an imperialist competitive algorithm with dynamic adjustment of parameters using fuzzy logic to adjust the Beta and Xi parameters. We are considering different fuzzy systems to measure the performance of the algorithm with six benchmark mathematical functions with different number of decades and performing 30 experiments for each case. The results demonstrate the efficiency of the fuzzy ICA algorithm in optimization problems and give us the guidelines for future work.

We describe in this paper the Bat Algorithm and a new proposed approach using interval type-2 fuzzy systems to dynamically adapt its parameters. The Bat Algorithm (denoted in the literature as BA) is a metaheuristic inspired in micro bats based on echolocation. We analyze in detail the behavior of this proposed modification using interval type-2 fuzzy logic and compare it with type-1 fuzzy logic to compare the performance of the proposed new algorithm based on the behavior of the mega bat.

In this paper, we present a new hybrid approach of flower pollination algorithm (FPA). This is a Bio-Inspired technique based on the pollination process carried out by the flowers. We used a Fuzzy inference system to adapt the probability of switching and this is the mechanism by which there is a change of global and local pollination; thus, the algorithm can explore and exploit in a different way to the original method. To validate in the best way the proposed method we present a comparison results among different optimization algorithms to evaluate the performance using a set of benchmark mathematical functions.

The main goal of this paper is to present a general study of the Grey Wolf Optimizer algorithm. We perform tests to determine in the first part which parameters are candidates to be dynamically adjusted and in the second stage to determine which are the parameters that have the greatest effect in the algorithm. We also present a justification and results of experiments as well as the benchmark functions that were used for the tests that are shown.

The contribution of this paper is to provide an analysis of the parameters of Gravitational Search Algorithm (GSA), to include a fuzzy logic system for dynamic parameter adaptation through the execution of the algorithm, in order to control the behavior of GSA based on some metrics like the iterations and the diversity of the agents in an specific moment of its execution.

This paper describes an optimization method based on particle swarm optimization (PSO) for ensemble neural networks with type-1 and type-2 fuzzy aggregation for forecasting complex time series. The time series that was considered in this paper to compare the hybrid approach with traditional methods is the Taiwan Stock Exchange (TAIEX), and the results shown are for the optimization of the structure of the ensemble neural network with type-1 and type-2 fuzzy integration. Simulation results show that ensemble approach produces good prediction of the Taiwan Stock Exchange.

According to the literature of particle swarm optimization (PSO), there are problems of local minima and premature convergence with this algorithm. A new algorithm is presented called the improved particle swarm optimization using the gradient descent method as operator of particle swarm incorporated into the Algorithm, as a function to test the improvement. The gradient descent method (BP Algorithm) helps not only to increase the global optimization ability, but also to avoid the premature convergence problem. The improved PSO algorithm IPSO is applied to Benchmark Functions. The results show that there is an improvement with respect to using the conventional PSO algorithm.

Eelgrass is a relevant sea grass species that provides important ecological services in near shore environments. The overall contribution of this species to human welfare is so important that upon threats to its permanence that associate to deleterious anthropogenic influences, a vigorous conservation effort has been recently enforced worldwide. Among restoration strategies transplanting plays a key role and the monitoring of the development of related plots is crucial to assess the restoration of the ecological features observed in donor populations. Since traditional eelgrass assessment methods are destructive their use in transplants could lead to undesirable effects such as alterations of shoot density and recruitment. Allometric methods can provide accurate proxies that sustain nondestructive estimations of variables required in the pertinent assessments. These constructs rely on extensive data sets for precise estimations of the involved parameters and also depend on precise estimations of the incumbent leaf area. The use of electronic scanning technologies for eelgrass leaf area estimation can enhance the nondestructive nature of associated allometric methods, because the necessary leaf area assessments could be obtained from digital images. But when a costly automatic leaf area meter is not available, we must rely on direct image processing, usually achieved through computationally costly Monte Carlo procedures. Previous results show that the amendment of simple genetic algorithms could drastically reduce the time required by regular Monte Carlo methods to achieve the estimation of the areas of individual eelgrass leaves. But even though this amendment, the completion of the task of measuring the areas of the leaves of a data set with an extension, as required for precise parameter estimation, still leads to a burdensome computational time. In this paper, we have explored the benefits that the addition of a master-slave parallel genetic algorithm to a Monte Carlo based estimation routine conveys in the aforementioned estimation task. We conclude that unless a suitable number of processors are involved, and also the proper mutation and crossover rates are contemplated the efficiency of the overall procedure will not be noticeably improved.

The aim of this study was to implement an optimal arrangement of equipment, instrumentation and medical personnel based on the weight and balance of the aircraft and to transfer humanitarian aid in a drone, by implementing artificial intelligence algorithms. This is due to the problems presented by the geographical layout of human settlements in southeast of the state of Chihuahua. The importance of this research is to understand the multivariable optimization associated with the path of a group of airplanes associated with different kinds of aerial in order to improve the evaluation of flooding and to send medical support and goods; to determine the optimal flight route, including speed, storage and travel resources. To determine the cost–benefit, this has been partnered with a travel plan to rescue people, which has as its principal basis the orography airstrip restriction, although this problem has been studied on several occasions by the literature failed to establish by supporting ubiquitous computing for interacting with the various values associated with the achievement of the group of drones and their cost–benefit of each issue of the company and comparing their individual trips for the rest of group. There are several factors that can influence in the achievement of a group of drones for our research. We propose the use of a bioinspired algorithm.

This work presents a high-performance implementation of a path planning algorithm based on parallel pseudo-bacterial potential field (parallel-PBPF) on a graphics processing unit (GPU) as an improvement to speed up the path planning computation in mobile robot navigation. Path planning is one of the most computationally intensive tasks in mobile robots and the challenge in dynamically changing environments. We show how data-intensive tasks in mobile robots can be processed efficiently through the use of GPUs. Experiments and simulation results are provided to show the effectiveness of the proposal.

Population pharmacokinetics (PopPK) models are used to characterize the behavior of a drug in a particular population. Construction of PopPK models requires the estimation of optimal PopPK parameters, which is a challenging task due to the characteristics of the PopPK database. Several estimation algorithms have been proposed for estimating PopPK parameters; however, the majority of these methods are based on maximum likelihood estimation methods that optimize the probability of observing data, given a model that requires the systematic computation of the first and second derivate of a multivariate likelihood function. This work presents a genetic algorithm for obtaining optimal PopPK parameters by directly optimizing the multivariate likelihood function avoiding the computation of the first and second derivate of the likelihood function.

This paper describes the optimization of a reactive controller system for a mobile autonomous robot using the CRA algorithm to adjust the parameters of each fuzzy controller. A comparison with the results obtained with genetic algorithms is also performed.

Edge detection is used on most pattern recognition algorithms for image processing, however, its main drawbacks are the detection of unreal edges and its computational cost; fuzzy edge detection is used to reduce false edges but at even higher computational cost. This paper presents a Field Programmable Gate Array (FPGA)-based hardware architecture that performs a real-time edge detection using fuzzy logic algorithms achieving a decrease in the amount of unreal edges detected while compensating the computational cost by using parallel and pipelining hardware design strategies. For image processing testing, image resolution is set to 480 × 640 pixels at 24 fps (frames per second), thus real-time processing requires 7,372,800 fuzzy logic inference per second (FLIPS). The proposed fuzzy logic edge detector is based on the morphological gradient; this algorithm performs the edge detection based in the gradient operator, getting vectors of edge direction, were the magnitude of these vectors determines if the pixel is edge or not. The hardware architecture processes each frame pixel by pixel with grayscale partial image inputs, at 8 bits resolution, represented with a 3 × 3 pixels matrix; subsequently the architecture executes the stages of the fuzzy logic system: fuzzification, inference, and defuzzification, however, taking advantage of the FPGAs versatility, the dedicated hardware-based processing is executed in parallel within a pipeline structure to achieve edge detection in real time. The real-time fuzzy edge detector is compared with several classic edge detectors to evaluate the performance in terms of quality of the edges and the processing rate in FLIPS.

A hybrid intelligent system is made of a powerful combination of soft computing techniques for reducing the complexity in solving difficult problems. Nowadays hypertension (high blood pressure) has a high prevalence in the world population and is the number one cause of mortality in Mexico, and this is why it is called a silent killer because it often has no symptoms. We design in this paper a hybrid model using modular neural networks, and as response integrator we use fuzzy systems to provide an accurate diagnosis of hypertension, so we can prevent future diseases in people based on the systolic pressure, diastolic pressure, and pulse of patients with ages between 15 and 95 years.

A study of the optimization of different types of membership functions (MF) using Bee Colony Optimization (BCO) for the stabilization of fuzzy controllers is presented. The main objective of the work is based on the main reasons for the comparative analysis of BCO as an optimization technique for the design of the Mamdani fuzzy controllers, specifically in tuning membership functions for two problems in fuzzy control. Simulations results confirmed that using the BCO to optimize the membership functions and the scaling gains of the fuzzy system improved the controller performance. The six metrics of the ITAE, ITSE, IAE, ISE, RMSE and MSE for the errors in control are implemented.

We propose a neuro-fuzzy hybrid model for the diagnosis of blood pressure to provide a diagnosis as accurate as possible based on intelligent computing techniques, such as neural networks and fuzzy logic. The neuro-fuzzy model uses a modular architecture which works with different number of layers and different learning parameters so that we can have a more accurate modeling. So for the better diagnosis and treatment of hypertension patients, an intelligent and accurate system is needed. In this study, we also design a fuzzy expert system to diagnose blood pressure for different patients. The fuzzy expert system is based on a set of inputs and rules. The input variables for this system are the systolic and diastolic pressures and the output variable is the blood pressures level. It is expected that this proposed neuro-fuzzy hybrid model can provide a faster, cheaper, and more accurate result.

In the early diagnosis of breast cancer, computer-aided diagnosis (CAD) systems help in the detection of abnormal tissue. Microcalcifications can be an early indication of breast cancer. This work describes the implementation of a new method for the detection of microcalcifications in mammographies. The images were obtained from the mini-MIAS database. In the proposed method, the images are preprocessed using an x and y gradient operators, the output of each filter is the input of a fuzzy system that will detect areas with high-tone variation. The next step consists of a cellular automaton that uses a set of local rules to eliminate noise and keep the pixels with higher probabilities of belonging to a microcalcification region. Comparative results are presented.

The tracking problem for an electrical actuator consisting of a DC motor and a reducer part (load) operating under uncertainty conditions due to friction and backlash is addressed. The Mamdani type fuzzy logic control will be designed to enforce the load position to track a prespecified reference trajectory. Since it is assumed that the dynamic model is not available, Lyapunov stability theory coupled together with the comparison principle will be used to conclude stability of the closed-loop system.

Differential evolution is a population-based metaheuristic that is widely used in Black-Box Optimization. The mutation is the main search operator and there are different implementation schemes reported in state of art literature. Nonetheless, such schemes lack mechanisms for an intensification stage, which can enable better search and avoid local optima. This article proposes a way to adapt the Covariance Matrix parameter of a Gaussian distribution that is used to generate a disturbance that improves the performance of two well-known mutation schemes. This disturbance allows working with problems with correlated variables. The test was performed over the CEC 2013 instances and the results were compared through the Friedman nonparametric test.

In this work a new bio-inspired metaheuristic based on the self-defense mechanism of plants is presented. This new optimization algorithm is applied to solve optimization problems, in this case optimization of mathematical functions for multiple variables, other works related to this, where the same algorithm is used with some modifications and improvements is presented in (Caraveo et al. in Advances in artificial intelligence and soft computing, Springer International Publishing, pp. 227–237, 2015 [3]). Since its inception the planet has gone through changes, so plants have had to adapt to these changes and adopt new techniques to defend from natural predators in this case. Many works have shown that plants have mechanisms of self-defense to protect themselves from predators. When the plants detect the presence of invading organisms this triggers a series of chemical reactions that are released to air and attract natural predators of the invading organism (Bennett and Wallsgrove in New Phytol, 127(4):617–633, 1994 [1]; Neyoy et al. Recent Advances on hybrid intelligent systems, Springer, Berlin, pp. 259–271, 2013 [10]; Ordeñana in Costa Rica, 63:22–32, 2002 [11]). For the development of this algorithm we consider as a main idea the predator prey model of Lotka and Volterra, where two populations are considered and the objective is to maintain a balance between the two populations.

The need to address more complex real-world problems gives rise to new research issues in many-objective optimization field. Recently, researchers have focused in developing algorithms able to solve optimization problems with more than three objectives known as many-objective optimization problems. Some methodologies have been developed into the context of this kind of problems, such as A2-NSGA-III that is an adaptive extension of the well-known NSGA-II (Non-dominated Sorting Genetic Algorithm II). A2-NSGA-III was developed for promoting a better spreading of the solutions in the Pareto front using an improved approach based on reference points. In this paper, a comparative study between NSGA-II and A2-NSGA-III is presented. We examine the performance of both algorithms by applying them to the project portfolio problem with 9 and 16 objectives. Our purpose is to validate the effectiveness of A2-NSGA-III to deal with many-objective problems and increase the variety of problems that this method can solve. Several quality indicators were used to measure the performance of the two algorithms.

In the recent years, Grammatical Evolution (GE) has been used as a representation of Genetic Programming (GP). GE can use a diversity of search strategies including Swarm Intelligence (SI). Bee Swarm Optimization (BSO) is part of SI and it tries to solve the main problems of the Particle Swarm Optimization (PSO): the premature convergence and the poor diversity. In this paper we propose using BSO as part of GE as strategies to generate heuristics that solve the Bin Packing Problem (BPP). A comparison between BSO, PSO, and BPP heuristics is performed through the nonparametric Friedman test. The main contribution of this paper is to propose a way to implement different algorithms as search strategy in GE. In this paper, it is proposed that the BSO obtains better results than the ones obtained by PSO, also there is a grammar proposed to generate online and offline heuristics to improve the heuristics generated by other grammars and humans.

Computing the Pathwidth of a graph is the problem of finding a linear ordering of the vertices such that the width of its corresponding path decomposition is minimized. This problem has been proven to be NP-hard. Currently, some of the best exact methods for generic graphs can be found in the mathematical software project called SageMath. This project provides an integer linear programming model (IPSAGE) and an enumerative algorithm (EASAGE), which is exponential in time and space. The algorithm EASAGE uses an array whose size grows exponentially with respect to the size of the problem. The purpose of this array is to improve the performance of the algorithm. In this chapter we propose two exact methods for computing pathwidth. More precisely, we propose a new integer linear programming formulation (IPPW) and a new enumerative algorithm (BBPW). The formulation IPPW generates a smaller number of variables and constraints than IPSAGE. The algorithm BBPW overcomes the exponential space requirement by using a last-in-first-out stack. The experimental results showed that, in average, IPPW reduced the number of variables by 33.3 % and the number of constraints by 64.3 % with respect to IPSAGE. This reduction of variables and constraints allowed IPPW to save approximately 14.9 % of the computing time of IPSAGE. The results also revealed that BBPW achieved a remarkable use of memory with respect to EASAGE. In average, BBPW required 2073 times less amount of memory than EASAGE for solving the same set of instances.

The course timetabling problem is one of the most difficult combinatorial problems, it requires the assignment of a fixed number of subjects into a number of time slots minimizing the number of student conflicts. This article presents a comparison between state-of-the-art hyper-heuristics and a newly proposed iterated variable neighborhood descent algorithm when solving the course timetabling problem. Our formulation can be seen as an adaptive iterated local search algorithm that combines several move operators in the improvement stage. Our improvement stage not only uses several neighborhoods, but it also incorporates state-of-the-art reinforcement learning mechanisms to adaptively select them on the fly. Our approach substitutes the adaptive improvement stage by a variable neighborhood descent (VND) algorithm. VND is an ingredient of the more general variable neighborhood search (VNS), a powerful metaheuristic that systematically exploits the idea of neighborhood change. This leads to a more effective search process according course timetabling benchmark results.

In this paper, the analytical parameter tuning for the Archive Multi-objective Simulated Annealing (AMOSA) is described. The analytical tuning method yields the initial and final temperature, and the maximum metropolis length. The analytically tuned AMOSA is used to solve the Heterogeneous Computing Scheduling Problem with independent tasks and it is compared versus the AMOSA without parameter tuning. We approach this problem as multi-objective, considering the makespan and the energy consumption. Also, in the last years this problem has gained importance due to the energy awareness in high performance computing centers (HPCC). The hypervolume, generational distance, and spread metrics were used in order to measure the performance of the implemented algorithms.

The aim of the Course Timetabling problem is to ensure that all the students take their required classes and adhere to resources that are available in the school. The set of constraints those must be considered in the design of timetabling involves students, teachers, and classrooms. In the state of the art are different methodologies of design for Course Timetabling problem, in this paper we extend the proposal from Soria in 2013, in which they consider variables of students and classrooms, with four set of generic structures. This paper uses Soria’s methodology to adding two more generic structures considering teacher restriction. We show an application of some different Metaheuristics using this methodology. Finally, we apply nonparametric test Wilcoxon signed-rank with the aim to find which metaheuristic algorithm shows a better performance in terms of quality.

The Cut width Minimization Problem is a NP-Hard problem that is found in the VLSI design, graph drawing, design of compilers and linguistics. Developing solutions that could solve it efficiently is important due to its impact in areas that are critical for society. It consists in finding the linear array of an undirected graph that minimizes the maximum number of edges that are cut. In this paper we propose a genetic algorithm applied to the Cut width Minimization Problem. As the configuration of a metaheuristic has a great impact on the performance, we also propose a Fuzzy Logic controller that is used to adjust the parameters of the GA during execution time to guide it during the exploration process.

Organizations are facing the problem of having more projects than resources to implement them. In this paper, we present a dialogue interaction module of a framework for a Decision Support System (DSS) to aid in the selection of public project portfolios. The Interaction module of this DSS is based on multiple argumentation schemes and dialogue games that not only allow the system to generate and justify a recommendation. This module is also able to obtain new information during the dialogue that allows changing the recommendation according to the Decision Maker’s preferences. Researchers have commonly addressed the public portfolio selection problem with multicriteria algorithms. However, in the real life the final selection of the solution depends on the decision maker (DM). We modeled the reasoning of DM by a Dialogue Corpus. This corpus is a database, supported by an argument tree that validates the system’s recommendations with the preferences of the DM.

The retrieval information models have been of important study since 1992. These models are based on comparing a user query and a collection of documents taking into account the concurrency of the terms, with the objective to classify a set of relevant documents and retrieve them to the user in accordance with the evaluations criterion. There are metrics to classify a set of documents according to the grade of similarity, such as cosine similarity and soft cosine measure. In this paper, we perform a comparative study of these similarity metrics. The Vector Space Model (VSM) was implemented for retrieving information. A sample of the Collection of the Association for Computing Machinery (CACM) in the domain of Computer Science was used in the evaluation. The experiment results show that the recall is of 96 % in both metrics, but the soft cosine achieves 2 % more in mean average precision.

Project portfolio selection is one of the most difficult, yet most important decision-making problems faced by many organizations in government and business sectors. The grey system theory proposed by Deng in 1982 is based on the assumption that a system is uncertain and that the information regarding the system is insufficient to build a relational analysis or to construct a model to characterize the system. The aim of this chapter is to compare a multi-attribute decision-making (MADM) that incorporates a system of preferences TOPSIS (Technique for Order Performance by Similarity to Ideal Solution) by Hwang and Yoon in (Multiple attribute decision making: methods and applications. Springer, Berlin, 1981) with TOPSIS-Grey by Lin et al. in (Expert Syst Appl 35:1638–1644, 2008).

Grammatical Evolution (GE) is a grammar-based form of Genetic Programming. In GE, a Mapping Process (MP) and a Backus–Naur Form grammar (defined in the problem context) are used to transform each individual’s genotype into its phenotype form (functional representation). There are several MPs proposed in the state-of-the-art, each of them defines how the individual’s genes are used to build its phenotype form. This paper compares two MPs: the Depth-First standard map and the Position Independent Grammatical Evolution (πGE). The comparison was performed using as use case the problem of the selection and generation of features for pattern recognition problems. A Wilcoxon Rank-Sum test was used to compare and validate the results of the different approaches.

Hyper-parameter tuning for support vector machines has been widely studied in the past decade. A variety of metaheuristics, such as Genetic Algorithms and Particle Swarm Optimization have been considered to accomplish this task. Notably, exhaustive strategies such as Grid Search or Random Search continue to be implemented for hyper-parameter tuning and have recently shown results comparable to sophisticated metaheuristics. The main reason for the success of exhaustive techniques is due to the fact that only two or three parameters need to be adjusted when working with support vector machines. In this chapter, we analyze two Estimation Distribution Algorithms, the Univariate Marginal Distribution Algorithm and the Boltzmann Univariate Marginal Distribution Algorithm, to verify if these algorithms preserve the effectiveness of Random Search and at the same time make more efficient the process of finding the optimal hyper-parameters without increasing the complexity of Random Search.

This study shows the results of a comparison of different measurement systems that help measure tweets virality within virtual communities. Likewise, the history of this type of virtual social networks in the context of marketing are essential to creating effective proposals for the study of computer systems, software developers and marketing professionals and advertising are presented. Ultimately, a proposal for a graphic tweets measurement system is presented.

Nowadays using a robust simulator is very important to support an organization in its first step to consolidate in the market, in this research we make a challenging organizational tool based on different components to make an adequate strategic planning methodology, unlike current applications, it is focused on an environment that goes beyond simple numerical forecasts and statistical processes. It is based on advanced components to optimize the strategies and stratagems to be followed within the company, helping businesses to achieve competitive advantage in the market. A business simulator is flexible, adaptive, has learning ability, is robust and fault tolerant. Our intelligent tool uses different methodologies to provide optimal strategies to improve competitiveness of a company, the ability of the model can provide strategies that are not obvious because they can find no obvious relationship among variables that can help the manager or leader of an organization to realize a better decision. This tool is an aid in the process of improving competitiveness because it supports the strategic decisions made in an organizational level.

In this paper, we propose the use of Non-Singleton Interval Type-2 Fuzzy Systems (NSIT2FI) automatically designed through genetic algorithms as integration method of modular neural networks (MNN’s) for multimodal biometrics. The goal is to obtain such fuzzy systems as integrators, better recognition rate, and best mean square error in MNN. The results shown comparison between interval type-2 fuzzy systems and Non-singleton Type-2 Fuzzy Systems, where we can observe showing a significant difference that we can get higher recognition rate using non-singleton type-2 fuzzy logic.