Review
Geosciences, Multidisciplinary
Richard M. Palin, M. Santosh
Summary: The theory of plate tectonics is widely accepted and provides a solid framework for describing and predicting the behavior of Earth's lithosphere. Interactions at the Earth's surface offer insight into the planet's inaccessible interior and allow speculation about geological characteristics of other rocky bodies.
Article
Automation & Control Systems
Imran Ahmed, Gwanggil Jeon, Francesco Piccialli
Summary: This article provides a comprehensive survey of AI and XAI-based methods in the context of Industry 4.0. It discusses the technologies enabling Industry 4.0, investigates the main methods used in the literature, and addresses the future research directions and the importance of responsible and human-centric AI and XAI systems.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Artificial Intelligence
Pasquale D'Alterio, Jonathan M. Garibaldi, Robert John, Christian Wagner
Summary: A novel type-reduction procedure for CIT2 fuzzy sets is introduced in this article based on the concept of switch indices. The algorithm is applied to a real-world classification problem and compared to other type-reduction approaches used in IT2 and CIT2 systems. The new algorithm is significantly faster than traditional methods while maintaining a high level of interpretability.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Regis Pierrard, Jean-Philippe Poli, Celine Hudelot
Summary: This article proposes an approach for explainable classification and annotation of images in high-stake applications, based on a transparent model and interpretable fuzzy relations. By setting expert knowledge in advance and extracting relevant relations using a fuzzy frequent itemset mining algorithm, rules for classification and constraints for annotation are successfully built. Experimental results demonstrate the approach's ability to successfully perform the target task and generate consistent and convincing explanations.
ARTIFICIAL INTELLIGENCE
(2021)
Article
Computer Science, Interdisciplinary Applications
Isaac Ronald Ward, Ling Wang, Juan Lu, Mohammed Bennamoun, Girish Dwivedi, Frank M. Sanfilippo
Summary: In this study, XAI techniques and Machine Learning models were used to predict ACS adverse outcomes based on individuals' health information, with rofecoxib and celecoxib drugs found to contribute to these predictions. The study achieved 72% accuracy in predicting ACS related adverse outcomes.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Article
Computer Science, Information Systems
Ziyi Kou, Daniel Zhang, Lanyu Shang, Dong Wang
Summary: Fauxtography is a type of posts containing misleading information on online social platforms. This paper focuses on explaining fauxtography posts by identifying the specific components that lead to the fauxtography. The authors present a Duo Explainable Fauxtography Detection Framework to generate explanations from both content and comment parts of fauxtography posts.
IEEE TRANSACTIONS ON BIG DATA
(2023)
Article
Computer Science, Information Systems
Gonzalo Napoles, Fabian Hoitsma, Andreas Knoben, Agnieszka Jastrzebska, Maikel Leon Espinosa
Summary: This paper presents a Prolog-based reasoning module for generating counterfactual explanations based on predictions from a black-box classifier. The approach involves four stages: preprocessing the dataset, transforming numerical features into symbolic ones using fuzzy clustering, encoding instances as Prolog rules, and computing the overall confidence of each rule using fuzzy-rough set theory. Additionally, a chatbot is implemented to handle natural language queries and generate counterfactual explanations.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Valerio La Gatta, Vincenzo Moscato, Marco Postiglione, Giancarlo Sperli
Summary: In this paper, a novel model-agnostic Explainable AI technique named CASTLE is proposed to provide rule-based explanations based on both the local and global model's workings. The framework has been evaluated on six datasets in terms of temporal efficiency, cluster quality and model significance, showing a 6% increase in interpretability compared to another state-of-the-art technique, Anchors.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Chemistry, Multidisciplinary
Ioannis D. Apostolopoulos, Peter P. Groumpos
Summary: Currently, the lack of explainability in artificial intelligence algorithms hinders their practical implementation and users' trust in the systems. This paper discusses the nature of fuzzy cognitive maps (FCMs) as a transparent and interpretable computational method in the field of explainable artificial intelligence (XAI). The study highlights the successful implementation of FCMs in various domains, such as medical decision-support systems, precision agriculture, and policy-making.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Artificial Intelligence
Carlo Combi, Beatrice Amico, Riccardo Bellazzi, Andreas Holzinger, Jason H. Moore, Marinka Zitnik, John H. Holmes
Summary: This paper focuses on the importance of explainable artificial intelligence (XAI) in the field of biomedicine. By bringing together researchers with different roles and perspectives, it explores XAI in depth and presents a series of requirements for achieving explainability in AI.
ARTIFICIAL INTELLIGENCE IN MEDICINE
(2022)
Article
Computer Science, Theory & Methods
Carmen Biedma-Rdguez, Maria Jose Gacto, Augusto Anguita-Ruiz, Rafael Alcala, Concepcion Maria Aguilera, Jesus Alcala-Fdez
Summary: Nowadays, there is an increasing demand for transparency in Artificial Intelligence models to understand their decision-making process and potential impact on human life and health. Fuzzy Rule-Based Classification Systems have been effectively used due to their easy interpretability. However, complex search spaces hinder the learning process and lead to complexity issues. Therefore, a fuzzy associative classification method is proposed to achieve a better balance between accuracy and complexity.
FUZZY SETS AND SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Francesco Piccialli, Vittorio Di Somma, Fabio Giampaolo, Salvatore Cuomo, Giancarlo Fortino
Summary: New technologies are revolutionizing medicine, with a key role played by data. Artificial intelligence, especially Deep Learning, is well-suited to handle the exponential growth of health-related information in the field of medicine, helping to build optimal neural networks for clinical problems as the amount of training data increases.
INFORMATION FUSION
(2021)
Review
Chemistry, Multidisciplinary
Manju Vallayil, Parma Nand, Wei Qi Yan, Hector Allende-Cid
Summary: This study explores the importance of explainability in the field of Automated Fact Verification (AFV) and highlights the current gaps and limitations. It finds that explainability in AFV lags behind the broader field of explainable AI (XAI). The study summarizes the elements of explainability in AFV, including architectural, methodological, and dataset-related aspects, and proposes possible recommendations for modifications to enhance the comprehensibility and acceptability of AI to the general society.
APPLIED SCIENCES-BASEL
(2023)
Article
Energy & Fuels
Christian Utama, Christian Meske, Johannes Schneider, Carolin Ulbrich
Summary: Efforts to support the energy transition and halt climate change have led to significant growth of renewable distributed generators, with photovoltaic systems being the fastest growing technology. However, high photovoltaic penetration in the electricity grid can lead to operational problems. This study proposes using artificial neural network to predict optimal reactive power dispatch in photovoltaic systems. The decentralized control leverages explainable artificial intelligence technique to identify non-local grid state measurements influencing system dispatch. Both centralized and decentralized controllers show superior performance, hindering voltage problems and line congestions while achieving energy savings compared to baseline strategies.
Article
Computer Science, Artificial Intelligence
Jerry M. Mendel, Piero P. Bonissone
Summary: This article discusses explainable artificial intelligence (XAI) for rule-based fuzzy systems, highlighting the importance of choosing antecedent membership function shapes for XAI. It provides a novel multi-step approach to obtain a simplified subset of rules, and offers a method to evaluate the quality of explanations.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2021)
Article
Automation & Control Systems
Miriam Seoane Santos, Pedro Henriques Abreu, Alberto Fernandez, Julian Luengo, Joao Santos
Summary: This work examines the impact of distance functions on K-Nearest Neighbours imputation of incomplete datasets. The experiments show that distance computation is significantly affected by missing data, and provide guidelines for selecting appropriate distance functions based on data characteristics and research objectives.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2022)
Article
Computer Science, Artificial Intelligence
Fatemeh Aghaeipoor, Mohammad Masoud Javidi, Alberto Fernandez
Summary: This article introduces an interpretable fuzzy classifier for Big Data, aiming to boost explainability by learning a compact yet accurate fuzzy model. Developed in a cell-based distributed framework, IFC-BD goes through three working stages of initial rule learning, rule generalization, and heuristic rule selection to move from a high number of specific rules to fewer, more general and confident rules. The proposed algorithm was found to improve the explainability and predictive performance of fuzzy rule-based classifiers in comparison to state-of-the-art approaches.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Alessio Bechini, Francesco Marcelloni, Alessandro Renda
Summary: This article presents a novel fuzzy clustering algorithm TSF-DBSCAN, which shows competitive performance in handling streaming data. The algorithm deals with outliers and evolution of data streams by introducing fuzziness and a fading model, while ensuring computational and memory efficiency.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2022)
Article
Computer Science, Information Systems
Alessio Bechini, Alessandro Bondielli, Jose Luis Corcuera Barcena, Pietro Ducange, Francesco Marcelloni, Alessandro Renda
Summary: In recent years, there has been a growing interest in profiling various aspects of city life, particularly in the context of smart cities. This interest has become even more relevant due to the realization that significant events, such as the Covid-19 pandemic, can have a profound impact on city life and lead to drastic changes. Identifying and analyzing these changes at both the city and neighborhood levels can be a crucial tool for effectively managing the current situation and developing strategies for future planning. The proposed framework in this article presents a novel methodology for modeling and tracking changes in city areas by extracting information from online newspaper articles.
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA
(2022)
Article
Computer Science, Artificial Intelligence
A. M. Garcia-Vico, C. J. Carmona, P. Gonzalez, M. J. del Jesus
Summary: This paper presents a distributed method based on evolutionary fuzzy systems for extracting and fusing patterns from data streams, and analyzes the adaptability and quality of the proposed method.
INFORMATION FUSION
(2023)
Article
Automation & Control Systems
Jesus Giraldez-Cru, Manuel Chica, Oscar Cordon
Summary: Consumers make decisions based on their perceptions of the market products using various decision-making strategies and heuristics. This study proposes the use of fuzzy linguistic information and four consumer DM heuristics to model qualitative consumer perceptions. Experimental analysis shows that combining these heuristics improves the model performance and provides a realistic representation of consumer perceptions.
INTERNATIONAL JOURNAL OF FUZZY SYSTEMS
(2023)
Article
Computer Science, Information Systems
Alessandro Renda, Pietro Ducange, Francesco Marcelloni, Dario Sabella, Miltiadis C. Filippou, Giovanni Nardini, Giovanni Stea, Antonio Virdis, Davide Micheli, Damiano Rapone, Leonardo Gomes Baltar
Summary: This article introduces the concept of federated learning (FL) of eXplainable Artificial Intelligence (XAI) models in advanced 5G towards 6G systems and discusses its applicability to the automated vehicle networking use case. It highlights the importance of considering XAI and FL for improving user experience and ensuring security and privacy in AI-based solutions for wireless network planning.
Editorial Material
Computer Science, Artificial Intelligence
Gustavo Olague, Mario Koppen, Oscar Cordon
Summary: This article introduces the field of Evolutionary Computer Vision (ECV), which is at the intersection of computer vision (CV) and evolutionary computation (EC). ECV utilizes evolutionary algorithms and metaheuristic approaches combined with analytical methods to achieve human-competitive results. It aims to design software and hardware solutions for challenging CV problems and enhance our understanding of visual processing in nature.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
J. Fumanal-Idocin, O. Cordon, H. Bustince
Summary: This work proposes a new algorithm called the Krypteia ensemble, which trains and optimizes a set of classifiers. The algorithm introduces diversity by generating different variations of the same task with distinct stochastic variables. It uses a set of agents to select individuals who excel in their assignments and minimizes redundancies in the resulting population. The study also explores stacking different Krypteia ensembles to build more complex classifiers and considers various aggregation functions for optimal performance.
INFORMATION FUSION
(2023)
Article
Engineering, Biomedical
Jose P. Amorim, Pedro H. Abreu, Alberto Fernandez, Mauricio Reyes, Joao Santos, Miguel H. Abreu
Summary: Healthcare agents are collecting large amounts of patient data, particularly in oncology. Decision-support systems based on deep learning techniques have been approved for clinical use, but their interpretability remains a barrier to their widespread adoption. This article aims to provide oncologists with a guide on how these methods make decisions and strategies to explain them. Theoretical concepts were illustrated using oncological examples and a literature review was conducted to identify research works in the field. The majority of studies focused on explaining the importance of tumor characteristics in predictions using multilayer perceptrons and convolutional neural networks. However, achieving interpretability while maintaining performance remains a significant challenge for artificial intelligence.
IEEE REVIEWS IN BIOMEDICAL ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Victor A. Vargas-Perez, Pablo Mesejo, Manuel Chica, Oscar Cordon
Summary: Agent-based models are used to simulate complex systems in marketing, and reinforcement learning is applied to train a brand agent to optimize marketing investment strategies. The learned strategy is dynamic and the agent makes investment decisions based on the current market state. The results show that the agent tends to invest in media channels with greater awareness impact while considering the situation and characteristics of the model instance.
INFORMATION FUSION
(2023)
Article
Biochemistry & Molecular Biology
Silvia D'Ambrosi, Stavros Giannoukakos, Mafalda Antunes-Ferreira, Carlos Pedraz-Valdunciel, Jillian W. P. Bracht, Nicolas Potie, Ana Gimenez-Capitan, Michael Hackenberg, Alberto Fernandez Hilario, Miguel A. Molina-Vila, Rafael Rosell, Thomas Wurdinger, Danijela Koppers-Lalic
Summary: This study investigates the synergistic contribution of circRNA and mRNA derived from blood platelets as biomarkers for lung cancer detection. Using a comprehensive bioinformatics pipeline, platelet-derived circRNA and mRNA from non-cancer individuals and lung cancer patients were analyzed. Machine learning algorithms were used to generate predictive classification models based on an optimal selected signature. The study demonstrates the potential of a multi-analyte-based approach using platelet-derived biomarkers for lung cancer detection.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2023)
Article
Automation & Control Systems
Guillermo Gomez-Trenado, Pablo Mesejo, Oscar Cordon
Summary: In this study, a method for predicting high-resolution cephalometric landmarks is proposed and evaluated. The results demonstrate that our approach outperforms other methods in a cephalometric landmarks dataset and achieves human-like performance in half of the cases.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Fatemeh Aghaeipoor, Mohammad Sabokrou, Alberto Fernandez
Summary: The explainability of deep neural networks has become a topic of interest, and this article proposes a fuzzy rule-based explainer system that helps understand the functioning of these networks. These systems maintain accuracy while reducing complexity and improving comprehensibility.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2023)
Article
Automation & Control Systems
Javier Fumanal-Idocin, Oscar Cordon, Gracaliz Pereira Dimuro, Antonio-Francisco Roldan Lopez-de-Hierro, Humberto Bustince
Summary: Social network analysis is a popular tool for studying relationships between interacting agents. However, it often overlooks domain-specific knowledge and its propagation. In this work, an extension of classical social network analysis is developed to incorporate external information. A new centrality measure, semantic value, and a new affinity function, semantic affinity, are proposed to establish fuzzy-like relationships among actors in the network. The method is applied to analyze different mythologies and compared with existing measures, showing more meaningful results.
IEEE TRANSACTIONS ON CYBERNETICS
(2023)