Computational Intelligence Methods and Applications – serie

After an introduction to renewable energy technologies, the authors present computational intelligence techniques for optimizing the manufacture of related technologies, including solar concentrators. In particular the authors present new applications for their neural classifiers for image and pattern recognition.The book will be of interest to researchers in computational intelligence, in particular in the domain of neural networks, and engineers engaged with renewable energy technologies.

This book describes how to use computational intelligence and artificial intelligence tools to improve the decision-making process in new product development. These approaches, including artificial neural networks and constraint satisfaction solutions, enable a more precise prediction of product development performance compared to widely used multiple regression models. They support decision-makers by providing more reliable information regarding, for example, project portfolio selection and project scheduling.The book is appropriate for computer scientists, management scientists, students and practitioners engaged with product innovation and computational intelligence applications.

This book describes how to use computational intelligence and artificial intelligence tools to improve the decision-making process in new product development. These approaches, including artificial neural networks and constraint satisfaction solutions, enable a more precise prediction of product development performance compared to widely used multiple regression models. They support decision-makers by providing more reliable information regarding, for example, project portfolio selection and project scheduling.The book is appropriate for computer scientists, management scientists, students and practitioners engaged with product innovation and computational intelligence applications.

This book focuses on the applied aspects of artificial intelligence using enterprise frameworks and technologies. The book is applied in nature and will equip the reader with the necessary skills and understanding for delivering enterprise ML technologies. It will be valuable for undergraduate and postgraduate students in subjects such as artificial intelligence and data science, and also for industrial practitioners engaged with data analytics and machine learning tasks. The book covers all of the key conceptual aspects of the field and provides a foundation for all interested parties to develop their own artificial intelligence applications.

This book focuses on the applied aspects of artificial intelligence using enterprise frameworks and technologies. The book is applied in nature and will equip the reader with the necessary skills and understanding for delivering enterprise ML technologies. It will be valuable for undergraduate and postgraduate students in subjects such as artificial intelligence and data science, and also for industrial practitioners engaged with data analytics and machine learning tasks. The book covers all of the key conceptual aspects of the field and provides a foundation for all interested parties to develop their own artificial intelligence applications.

This book describes a general hybrid metaheuristic for combinatorial optimization labeled Construct, Merge, Solve & Adapt (CMSA). The general idea of standard CMSA is the following one. At each iteration, a number of valid solutions to the tackled problem instance are generated in a probabilistic way. Hereby, each of these solutions is composed of a set of solution components. The components found in the generated solutions are then added to an initially empty sub-instance. Next, an exact solver is applied in order to compute the best solution of the sub-instance, which is then used to update the sub-instance provided as input for the next iteration. In this way, the power of exact solvers can be exploited for solving problem instances much too large for a standalone application of the solver.Important research lines on CMSA from recent years are covered in this book. After an introductory chapter about standard CMSA, subsequent chapters cover a self-adaptive CMSA variant as well as a variant equipped with a learning component for improving the quality of the generated solutions over time. Furthermore, on outlining the advantages of using set-covering-based integer linear programming models for sub-instance solving, the author shows how to apply CMSA to problems naturally modelled by non-binary integer linear programming models. The book concludes with a chapter on topics such as the development of a problem-agnostic CMSA and the relation between large neighborhood search and CMSA. Combinatorial optimization problems used in the book as test cases include the minimum dominating set problem, the variable-sized bin packing problem, and an electric vehicle routing problem.The book will be valuable and is intended for researchers, professionals and graduate students working in a wide range of fields, such as combinatorial optimization, algorithmics, metaheuristics, mathematical modeling, evolutionary computing, operations research, artificial intelligence, or statistics.

This book describes a general hybrid metaheuristic for combinatorial optimization labeled Construct, Merge, Solve & Adapt (CMSA). The general idea of standard CMSA is the following one. At each iteration, a number of valid solutions to the tackled problem instance are generated in a probabilistic way. Hereby, each of these solutions is composed of a set of solution components. The components found in the generated solutions are then added to an initially empty sub-instance. Next, an exact solver is applied in order to compute the best solution of the sub-instance, which is then used to update the sub-instance provided as input for the next iteration. In this way, the power of exact solvers can be exploited for solving problem instances much too large for a standalone application of the solver.Important research lines on CMSA from recent years are covered in this book. After an introductory chapter about standard CMSA, subsequent chapters cover a self-adaptive CMSA variant as well as a variant equipped with a learning component for improving the quality of the generated solutions over time. Furthermore, on outlining the advantages of using set-covering-based integer linear programming models for sub-instance solving, the author shows how to apply CMSA to problems naturally modelled by non-binary integer linear programming models. The book concludes with a chapter on topics such as the development of a problem-agnostic CMSA and the relation between large neighborhood search and CMSA. Combinatorial optimization problems used in the book as test cases include the minimum dominating set problem, the variable-sized bin packing problem, and an electric vehicle routing problem.The book will be valuable and is intended for researchers, professionals and graduate students working in a wide range of fields, such as combinatorial optimization, algorithmics, metaheuristics, mathematical modeling, evolutionary computing, operations research, artificial intelligence, or statistics.

This book explains efficient solutions for segmenting the intensity levels of different types of multilevel images. The authors present hybrid soft computing techniques, which have advantages over conventional soft computing solutions as they incorporate data heterogeneity into the clustering/segmentation procedures.This is a useful introduction and reference for researchers and graduate students of computer science and electronics engineering, particularly in the domains of image processing and computational intelligence.

This book explains efficient solutions for segmenting the intensity levels of different types of multilevel images. The authors present hybrid soft computing techniques, which have advantages over conventional soft computing solutions as they incorporate data heterogeneity into the clustering/segmentation procedures.This is a useful introduction and reference for researchers and graduate students of computer science and electronics engineering, particularly in the domains of image processing and computational intelligence.

Covering a wide range of notions concerning hesitant fuzzy set and its extensions, this book provides a comprehensive reference to the topic. In the case where different sources of vagueness appear simultaneously, the concept of fuzzy set is not able to properly model the uncertainty, imprecise and vague information. In order to overcome such a limitation, different types of fuzzy extension have been introduced so far. Among them, hesitant fuzzy set was first introduced in 2010, and the existing extensions of hesitant fuzzy set have been encountering an increasing interest and attracting more and more attentions up to now. It is not an exaggeration to say that the recent decade has seen the blossoming of a larger set of techniques and theoretical outcomes for hesitant fuzzy set together with its extensions as well as applications.As the research has moved beyond its infancy, and now it is entering a maturing phase with increased numbers and types of extensions, this book aims to give a comprehensive review of such researches. Presenting the review of many and important types of hesitant fuzzy extensions, and including references to a large number of related publications, this book will serve as a useful reference book for researchers in this field.

Covering a wide range of notions concerning hesitant fuzzy set and its extensions, this book provides a comprehensive reference to the topic. In the case where different sources of vagueness appear simultaneously, the concept of fuzzy set is not able to properly model the uncertainty, imprecise and vague information. In order to overcome such a limitation, different types of fuzzy extension have been introduced so far. Among them, hesitant fuzzy set was first introduced in 2010, and the existing extensions of hesitant fuzzy set have been encountering an increasing interest and attracting more and more attentions up to now. It is not an exaggeration to say that the recent decade has seen the blossoming of a larger set of techniques and theoretical outcomes for hesitant fuzzy set together with its extensions as well as applications.As the research has moved beyond its infancy, and now it is entering a maturing phase with increased numbers and types of extensions, this book aims to give a comprehensive review of such researches. Presenting the review of many and important types of hesitant fuzzy extensions, and including references to a large number of related publications, this book will serve as a useful reference book for researchers in this field.

Using metaheuristics to enhance machine learning techniques has become trendy and has achieved major successes in both supervised (classification and regression) and unsupervised (clustering and rule mining) problems. Furthermore, automatically generating programs via metaheuristics, as a form of evolutionary computation and swarm intelligence, has now gained widespread popularity. This book investigates different ways of integrating metaheuristics into machine learning techniques, from both theoretical and practical standpoints. It explores how metaheuristics can be adapted in order to enhance machine learning tools and presents an overview of the main metaheuristic programming methods. Moreover, real-world applications are provided for illustration, e.g., in clustering, big data, machine health monitoring, underwater sonar targets, and banking.

Using metaheuristics to enhance machine learning techniques has become trendy and has achieved major successes in both supervised (classification and regression) and unsupervised (clustering and rule mining) problems. Furthermore, automatically generating programs via metaheuristics, as a form of evolutionary computation and swarm intelligence, has now gained widespread popularity. This book investigates different ways of integrating metaheuristics into machine learning techniques, from both theoretical and practical standpoints. It explores how metaheuristics can be adapted in order to enhance machine learning tools and presents an overview of the main metaheuristic programming methods. Moreover, real-world applications are provided for illustration, e.g., in clustering, big data, machine health monitoring, underwater sonar targets, and banking.

With the rapid advancement of sensor technology and digital system, the capabilities of network communication have significantly improved, allowing multiple computing nodes to exchange information and collaborate seamlessly through networks. This progress has accelerated the development of distributed optimization theory and its applications in emerging fields such as low altitude economy, big data, and artificial intelligence. These emerging domains usually involve solving complex large-scale optimization problems, making it difficult for traditional centralized methods to handle. Therefore, it is necessary to study distributed algorithms to solve complex optimization problems in large-scale networked systems. In addition, the emergence of applications of large language model further stimulates researchers’ growing interest in distributed optimization. This book provides the advanced methods and techniques of distributed optimization in networked systems, and thus is necessary and important for the research community. This book focuses on designing high-performance algorithms for solving more practical and complex optimization problems (multi-block optimization, composite optimization, constrained optimization, optimization with diversity objective functions, etc.) in the context of distributed optimization in networked systems and their successful application to real-world applications (model predictive control, smart grids, etc.). Readers may be particularly interested in the book on consensus and optimization protocols, forward-backward splitting methods, proximal gradient methods, primal-dual methods, fixed point methods, asynchronous communication/computaion mechanisms, randomized block coordinate techniques, operator splitting schemes, uncoordinated step sizes strategies, etc., in the process of distributed optimization in various networked systems. This book will introduce readers to the latest and advanced techniques in “Network-System Research and Distributed Composite Algorithm Design”, and help them develop their own novel distributed algorithms that have practical applications. The prerequisite for understanding this book is to master basic mathematical knowledge, including graph theory, matrix theory, linear algebra, probability theory, etc. This book is meant for the researcher and engineer who uses distributed optimization algorithms in fields like control theory, electronic information, artificial intelligence, and computer science, etc. It can also serve as complementary reading for distributed optimization in networked systems at the post-graduate level.

In manufacturing contexts, Industry 4.0 and Industry 5.0 are forefront names for the concepts, technologies, practices, and skills underlying smart factories and industries of the future. Industry 5.0 extends the technological pillars of Industry 4.0 by putting more emphasis on human centricity, sustainability and resilience in the engineering (architecture, design and development), deployment (strategic/technology road mapping, business/digital transformation, and operationalization), and management (operational excellence and continual improvement) of complex cyber-physical production systems.This book contributes to bridging some key research gaps that hinder a full adoption of these trends in industry. Key topics focus on how to ensure applicability of human centricity, sustainability and resilience requirements in all three aspects related to system design, transformation, and operation, namely (1) human behaviors, skills, and competencies; (2) activities and business processes; and (3) system structure (including technology and information), behavior (including planning, monitoring, and control), and outcomes (including products and services). Case studies are provided in various chapters of the book to illustrate these key topics.

In today’s world, the concept of a circular economy has emerged as a vital solution to address pressing environmental concerns and foster sustainable development. At its core, the circular economy aims to minimize waste generation by keeping resources in use for as long as possible, thereby reducing the strain on natural ecosystems and promoting resource efficiency.Operations research and data science play a crucial role in implementing circular economy principles by bridging the gap between data generated from new technologies and decision-making processes.This book serves as a comprehensive guide, illustrating how operations research and data science techniques can function as a decision support system, facilitating the optimal implementation of circular economy principles. With a particular focus on waste management, the book delves into various aspects such as recycling, reusing, and refurbishing, providing practical real-world applications and examples.By blending theoretical frameworks with practical insights, the book empowers both professionals and researchers with the essential tools and knowledge to navigate the intricacies of sustainable waste management. It underscores the significance of operations research and data science in facilitating the optimal implementation of circular economy principles. Offering comprehensive insights and actionable examples, readers acquire valuable guidance for implementing positive changes in waste management practices.

The escalating tide of waste and environmental threats demands a paradigm shift in how we produce and consume waste. The circular economy emerges as a beacon of hope, offering a transformative approach that breaks free from the traditional "take make dispose" model. This innovative system prioritizes sustainable practices throughout a product's lifecycle, promoting reuse, recycling, and regeneration. By fostering the development of robust and closed loop supply chains, the circular economy ensures that resources are kept in use for as long as possible, thus minimizing waste and environmental impact.To achieve these ambitious goals, the circular economy leverages cutting edge technologies such as operations research and data science. These tools empower businesses by constructing decision support systems that effectively handle complex environmental and economic constraints. Operational research helps in optimizing resource allocation, production processes, and logistics, while data science provides insights through data analysis and predictive modeling.This book aims to detail the application of operations research and data science methods as decision support systems to manage complex constraints. This book presents the most recent advancements and real-world applications of these methods within the framework of supply chain management in the circular economy. Additionally, it explores case studies that highlight successful implementations and the benefits achieved. The book offers comprehensive insights and practical applications, serving as a valuable guide for students, researchers, and professionals looking to drive sustainability in their organizations and communities.

This book draws on a range of intelligent computing methodologies to effectively detect and classify various carcinogenic diseases. These methodologies, which have been developed on a sound foundation of gene-level, cell-level and tissue-level carcinogenic datasets, are discussed in Chapters 1 and 2.  Chapters 3, 4 and 5 elaborate on several intelligent gene selection methodologies such as filter methodologies and wrapper methodologies. In addition, various gene selection philosophies for identifying relevant carcinogenic genes are described in detail. In turn, Chapters 6 and 7 tackle the issues of using cell-level and tissue-level datasets to effectively detect carcinogenic diseases. The performance of different intelligent feature selection techniques is evaluated on cell-level and tissue-level datasets to validate their effectiveness in the context of carcinogenic disease detection.  In closing, the book presents illustrative case studies that demonstrate the value of intelligent computing strategies.

This book draws on a range of intelligent computing methodologies to effectively detect and classify various carcinogenic diseases. These methodologies, which have been developed on a sound foundation of gene-level, cell-level and tissue-level carcinogenic datasets, are discussed in Chapters 1 and 2.  Chapters 3, 4 and 5 elaborate on several intelligent gene selection methodologies such as filter methodologies and wrapper methodologies. In addition, various gene selection philosophies for identifying relevant carcinogenic genes are described in detail. In turn, Chapters 6 and 7 tackle the issues of using cell-level and tissue-level datasets to effectively detect carcinogenic diseases. The performance of different intelligent feature selection techniques is evaluated on cell-level and tissue-level datasets to validate their effectiveness in the context of carcinogenic disease detection.  In closing, the book presents illustrative case studies that demonstrate the value of intelligent computing strategies.

This book provides a comprehensive review of the latest research in the area of explainable artificial intelligence (XAI) in health informatics. It focuses on how explainable AI models can work together with humans to assist them in decision-making, leading to improved diagnosis and prognosis in healthcare. This book includes a collection of techniques and systems of XAI in health informatics and gives a wider perspective about the impact created by them. The book covers the different aspects, such as robotics, informatics, drugs, patients, etc., related to XAI in healthcare.The book is suitable for both beginners and advanced AI practitioners, including students, academicians, researchers, and industry professionals. It serves as an excellent reference for undergraduate and graduate-level courses on AI for medicine/healthcare or XAI for medicine/healthcare. Medical institutions can also utilize this book as reference material and provide tutorials to medical professionals on how the XAI techniques can contribute to trustworthy diagnosis and prediction of the diseases.

This book provides a comprehensive review of the latest research in the area of explainable artificial intelligence (XAI) in health informatics. It focuses on how explainable AI models can work together with humans to assist them in decision-making, leading to improved diagnosis and prognosis in healthcare. This book includes a collection of techniques and systems of XAI in health informatics and gives a wider perspective about the impact created by them. The book covers the different aspects, such as robotics, informatics, drugs, patients, etc., related to XAI in healthcare.The book is suitable for both beginners and advanced AI practitioners, including students, academicians, researchers, and industry professionals. It serves as an excellent reference for undergraduate and graduate-level courses on AI for medicine/healthcare or XAI for medicine/healthcare. Medical institutions can also utilize this book as reference material and provide tutorials to medical professionals on how the XAI techniques can contribute to trustworthy diagnosis and prediction of the diseases.

Membrane computing is a class of distributed and parallel computing models inspired by living cells. Spiking neural P systems are neural-like membrane computing models, representing an interdisciplinary field between membrane computing and artificial neural networks, and are considered one of the third-generation neural networks. Models and applications constitute two major research topics in spiking neural P systems. The entire book comprises two parts: models and applications. In the model part, several variants of spiking neural P systems and fuzzy spiking neural P systems are introduced. Subsequently, their computational completeness is discussed, encompassing digital generation/accepting devices, function computing devices, and language generation devices. This discussion is advantageous for researchers in the fields of membrane computing, biologically inspired computing, and theoretical computer science, aiding in understanding the distributed computing model of spiking neural P systems. In the application part, the application of spiking neural P systems in time series prediction, image processing, sentiment analysis, and fault diagnosis is examined. This offers a novel method and model for researchers in artificial intelligence, data mining, image processing, natural language processing, and power systems. Simultaneously, it furnishes engineering and technical personnel in these fields with a powerful, efficient, reliable, and user-friendly set of tools and methods.

Membrane computing is a class of distributed and parallel computing models inspired by living cells. Spiking neural P systems are neural-like membrane computing models, representing an interdisciplinary field between membrane computing and artificial neural networks, and are considered one of the third-generation neural networks. Models and applications constitute two major research topics in spiking neural P systems. The entire book comprises two parts: models and applications. In the model part, several variants of spiking neural P systems and fuzzy spiking neural P systems are introduced. Subsequently, their computational completeness is discussed, encompassing digital generation/accepting devices, function computing devices, and language generation devices. This discussion is advantageous for researchers in the fields of membrane computing, biologically inspired computing, and theoretical computer science, aiding in understanding the distributed computing model of spiking neural P systems. In the application part, the application of spiking neural P systems in time series prediction, image processing, sentiment analysis, and fault diagnosis is examined. This offers a novel method and model for researchers in artificial intelligence, data mining, image processing, natural language processing, and power systems. Simultaneously, it furnishes engineering and technical personnel in these fields with a powerful, efficient, reliable, and user-friendly set of tools and methods.

Deep learning has achieved impressive results in image classification, computer vision and natural language processing. To achieve better performance, deeper and wider networks have been designed, which increase the demand for computational resources. The number of floating-point operations (FLOPs) has increased dramatically with larger networks, and this has become an obstacle for convolutional neural networks (CNNs) being developed for mobile and embedded devices. In this context, our book will focus on CNN compression and acceleration, which are important for the research community. We will describe numerous methods, including parameter quantization, network pruning, low-rank decomposition and knowledge distillation. More recently, to reduce the burden of handcrafted architecture design, neural architecture search (NAS) has been used to automatically build neural networks by searching over a vast architecture space. Our book will also introduce NAS due to its superiority and state-of-the-art performance in various applications, such as image classification and object detection. We also describe extensive applications of compressed deep models on image classification, speech recognition, object detection and tracking. These topics can help researchers better understand the usefulness and the potential of network compression on practical applications. Moreover, interested readers should have basic knowledge about machine learning and deep learning to better understand the methods described in this book.

Deep learning has achieved impressive results in image classification, computer vision and natural language processing. To achieve better performance, deeper and wider networks have been designed, which increase the demand for computational resources. The number of floating-point operations (FLOPs) has increased dramatically with larger networks, and this has become an obstacle for convolutional neural networks (CNNs) being developed for mobile and embedded devices. In this context, our book will focus on CNN compression and acceleration, which are important for the research community. We will describe numerous methods, including parameter quantization, network pruning, low-rank decomposition and knowledge distillation. More recently, to reduce the burden of handcrafted architecture design, neural architecture search (NAS) has been used to automatically build neural networks by searching over a vast architecture space. Our book will also introduce NAS due to its superiority and state-of-the-art performance in various applications, such as image classification and object detection. We also describe extensive applications of compressed deep models on image classification, speech recognition, object detection and tracking. These topics can help researchers better understand the usefulness and the potential of network compression on practical applications. Moreover, interested readers should have basic knowledge about machine learning and deep learning to better understand the methods described in this book.

This book delves into practical implementation of evolutionary and metaheuristic algorithms to advance the capacity of machine learning. The readers can gain insight into the capabilities of data-driven evolutionary optimization in materials mechanics, and optimize your learning algorithms for maximum efficiency. Or unlock the strategies behind hyperparameter optimization to enhance your transfer learning algorithms, yielding remarkable outcomes. Or embark on an illuminating journey through evolutionary techniques designed for constructing deep-learning frameworks. The book also introduces an intelligent RPL attack detection system tailored for IoT networks. Explore a promising avenue of optimization by fusing Particle Swarm Optimization with Reinforcement Learning. It uncovers the indispensable role of metaheuristics in supervised machine learning algorithms. Ultimately, this book bridges the realms of evolutionary dynamic optimization andmachine learning, paving the way for pioneering innovations in the field.