Article
Computer Science, Artificial Intelligence
Yeonggyu Yun, Hye-Young Jung
Summary: The study examines the effects of policy reforms on public medical insurance on households using fuzzy cognitive map. A hybrid approach is adopted to construct maps for low-income households and general households separately. Results show that government subsidy increases have the largest impacts on households, demonstrating the flexibility and extensibility of FCM.
Article
Automation & Control Systems
Hongling Qiu, Heng Liu, Xiulan Zhang
Summary: This paper proposes a composite learning adaptive backstepping fuzzy control method using historical data to improve the approximation accuracy of fuzzy logic systems for functional uncertainties of fractional-order nonlinear systems. The method employs a command filter to resolve the complexity issue caused by differentiating virtual controllers repeatedly, and defines a compensation signal to reduce the impact of filtered errors on control performance. A modified prediction error is also introduced to construct a composite parameter adaptation law by integrating the compensated error dynamic system equation. The method ensures the boundedness of all signals in the closed-loop system and achieves precise approximation to functional uncertainties. Numerical simulation examples demonstrate the superiority of the proposed method.
INTERNATIONAL JOURNAL OF FUZZY SYSTEMS
(2023)
Article
Green & Sustainable Science & Technology
Oguz Emir, Sule Onsel Ekici
Summary: In recent years, waste management has gained attention due to sustainability concerns and the depletion of natural resources. Food waste management is particularly important given the growing population and hunger crisis. Integrated assessment models (IAMs) have been commonly used to study food waste and provide insights to policymakers, while the Fuzzy Cognitive Map (FCM) extended with intuitionistic fuzzy sets offers a framework for analyzing interactions between factors and prioritizing policies.
JOURNAL OF CLEANER PRODUCTION
(2023)
Article
Computer Science, Cybernetics
Pinar Kocabey Ciftci, Zeynep Didem Unutmaz Durmusoglu
Summary: This article proposes a novel hybrid simulation model that combines agent-based model (ABM) and multistage learning-based fuzzy cognitive map (FCM) to understand complex tobacco use behavior. The ABM represents individual level behaviors while the FCM serves as a decision support mechanism. The model considers socio-demographic characteristics, tobacco control policies, and social network effects to reflect the tobacco use system in Turkey, and analyzes the impact of plain package and COVID-19 under different scenarios.
Article
Computer Science, Information Systems
Bicheng Yan, Xiaoqiang Jiang, Khalid A. Alattas, Chunwei Zhang, Ardashir Mohammadzadeh
Summary: This paper presents a novel fuzzy control strategy for generating limit cycles with specific behaviors in nonlinear complex dynamics. The proposed controller utilizes interval type-3 fuzzy logic, enhancing the quality of the closed-loop response and robust performance. An adaptively learned backstepping controller based on fuzzy control is employed to analyze convergence and robustness. Various simulations are conducted to validate the effectiveness of the fuzzy-based control law and adaptation rules.
Article
Chemistry, Physical
Henryk Otwinowski, Jaroslaw Krzywanski, Dariusz Urbaniak, Tomasz Wylecial, Marcin Sosnowski
Summary: This paper introduces a knowledge-based air classification system for efficient material classification. The model considers various operating parameters and has been successfully validated against experimental data. The research in this paper is of great importance for optimizing the classification process.
Article
Neurosciences
M. Ganesh Kumar, Cheston Tan, Camilo Libedinsky, Shih-Cheng Yen, Andrew Y. Y. Tan
Summary: The study shows that classic agents can learn to navigate to a single reward location and adapt to reward location displacement, but are unable to learn multiple cue-reward location tasks. By improving the agent's architecture and learning methods, this limitation can be overcome.
Article
Automation & Control Systems
Ahmad M. El-Nagar, Ahmad M. Zaki, F. A. S. Soliman, Mohammad El-Bardini
Summary: A hybrid deep learning neural network controller (HDLNNC) for nonlinear systems is proposed in this paper, and experimental results show that the proposed controller can improve system performance compared to other controllers.
INTERNATIONAL JOURNAL OF CONTROL
(2023)
Article
Automation & Control Systems
Rogerio P. Pereira, Eduardo J. F. Andrade, Jose L. F. Salles, Carlos T. Valadao, Ravena S. Monteiro, Gustavo Maia de Almeida, Marco A. S. L. Cuadros, Teodiano F. Bastos-Filho
Summary: This paper proposes two learning-type regulatory controllers, which use fuzzy logic and adaptive techniques to counteract cyclical disturbances in industrial processes and automatically adjust the controllers to reduce output oscillations. Numerical results verify the effectiveness of these two controllers.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Xia Wang, Bin Xu, Shuai Li, Qinmin Yang, Quanyong Fan
Summary: This study investigates a composite learning fuzzy control for a class of stochastic nonlinear strict-feedback systems, emphasizing the accuracy of fuzzy learning. By constructing a serial-parallel estimation model, a more accurate feedback information composite fuzzy updating law is designed. Through simulation tests, it is proved that the proposed scheme can effectively solve system uncertainty.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Ankita Dhar, Himadri Mukherjee, Niladri Sekhar Dash, Kaushik Roy
Summary: Automatic text categorization involves sorting text documents into predefined categories using machine learning algorithms. Key factors in text categorization include word sense, semantic relationships, and different methods within the conventional approaches.
ARTIFICIAL INTELLIGENCE REVIEW
(2021)
Article
Computer Science, Software Engineering
Yasin Turkyilmaz, Arafat Senturk, Muhammed Enes Bayrakdar
Summary: In this article, a machine learning based malicious signal detection system is proposed for cognitive radio networks. The system utilizes fuzzy logic for the security categorization of spectrum sensing signals and is validated with the results obtained from a fuzzy logic based approach. The random forest method performs the best among all machine learning methods for signal detection.
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2023)
Article
Computer Science, Interdisciplinary Applications
Mustafa Jahangoshai Rezaee, Samuel Yousefi, Majid Baghery, Ripon K. Chakrabortty
Summary: This study investigates the cause-and-effect relationships between goals and performance measurements in organizations' Balanced Scorecard using the Fuzzy Cognitive Map method, aiming to optimize the selection and evaluation of improvement projects. By calculating the effectiveness value of improvement projects on organization goals and prioritizing based on resource constraints using fuzzy SBDEA, a unique decision support system tool is provided for decision makers.
COMPUTERS & INDUSTRIAL ENGINEERING
(2021)
Article
Automation & Control Systems
Maxim Bobyr, Alexander Arkhipov, Sergey Emelyanov, Natalya Milostnaya
Summary: This article presents a new fuzzy method for creating a depth map, which combines Canny detector with a three-level fuzzy system to improve accuracy. The method includes three fuzzy models that eliminate errors, white color artifacts, and optimize the depth map. The study found that the labels of membership functions and the combination of t and s-norms significantly affect the accuracy of the resulting depth map.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Libin Wang, Huanqing Wang, Peter Xiaoping Liu, Song Ling, Siwen Liu
Summary: This article presents a fuzzy finite-time command filtering output feedback control method for a class of nonlinear systems. The method solves the computational complexity problem and ensures the finite-time boundedness of signals and convergence of tracking error by introducing fuzzy logic system and fuzzy state observer.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2022)
Article
Automation & Control Systems
Salah Hasan Al-subhi, Elpiniki I. Papageorgiou, Pedro Pinero Perez, Gaafar Sadeq S. Mahdi, Luis Alvarado Acuna
Summary: Fuzzy cognitive maps are widely used in modeling large and complex systems, but they have limitations in representing uncertainties. Neutrosophic cognitive maps are proposed as an extension to address this issue, but they still lack the ability to quantify uncertainty. In addition, some decision-making problems involve interrelated decisions that are difficult to prioritize in sequential order.
INTERNATIONAL JOURNAL OF FUZZY SYSTEMS
(2021)
Article
Chemistry, Multidisciplinary
Nikolaos Papandrianos, Elpiniki Papageorgiou
Summary: This research paper investigates the automatic diagnosis of ischemia or infarction in coronary artery disease (CAD) patients using deep learning and convolutional neural networks. The study introduces a robust CNN model for myocardial perfusion image classification and achieves high accuracy in differentiating between healthy and affected patients. The results demonstrate the effectiveness of the proposed deep learning approaches for CAD diagnosis using SPECT MPI scans.
APPLIED SCIENCES-BASEL
(2021)
Article
Computer Science, Artificial Intelligence
Gayathri Soman, M. V. Vivek, M. V. Judy, Elpiniki Papageorgiou, Vassilis C. Gerogiannis
Summary: This paper focuses on emotion recognition and proposes a distributed ensemble model for emotion classification. The results show that the proposed ensemble model achieves high accuracy in emotion classification.
Article
Energy & Fuels
Athanasios Anagnostis, Serafeim Moustakidis, Elpiniki Papageorgiou, Dionysis Bochtis
Summary: This study presents an alternative data-driven approach for modelling the temperature dynamics of thermal energy storage (TES) systems. A hybrid bimodal LSTM architecture is proposed to model the temperature dynamics of different TES components, and a cascading modelling framework is introduced to integrate the results of the individual modelling components. Experimental analysis demonstrates the low-error performance and real-time deployability of the proposed approach, as well as the effectiveness of the integrated energy framework operating in fine timescales.
Article
Computer Science, Artificial Intelligence
Serafeim Moustakidis, Athanasios Siouras, Konstantinos Vassis, Ioannis Misiris, Elpiniki Papageorgiou, Dimitrios Tsaopoulos
Summary: This study aims to identify risk factors for injuries in CrossFit and develop a machine learning model for predicting injuries. Through a survey and machine learning process conducted in Greece, the researchers successfully built a predictive model and achieved an AUC of 77.93% on the selected risk factors.
Article
Chemistry, Multidisciplinary
Nikolaos Papandrianos, Anna Feleki, Serafeim Moustakidis, Elpiniki Papageorgiou, Ioannis D. Apostolopoulos, Dimitris J. Apostolopoulos
Summary: The study introduces an explainable deep learning methodology for automatic classification of coronary artery disease. The model achieves 93.3% accuracy and 94.58% AUC in identifying CAD status, demonstrating efficient performance and stability.
APPLIED SCIENCES-BASEL
(2022)
Article
Medicine, General & Internal
Nikolaos Papandrianos, Anna Feleki, Elpiniki Papageorgiou, Chiara Martini
Summary: This research utilizes deep learning models to classify heart images and can assist in nuclear medicine and clinical decision-making, improving the accuracy of diagnosis for coronary artery disease.
JOURNAL OF CLINICAL MEDICINE
(2022)
Article
Energy & Fuels
Katarzyna Poczeta, Elpiniki Papageorgiou
Summary: This paper presents a novel approach to energy use forecasting, utilizing a nested fuzzy cognitive map. The experiments demonstrate the effectiveness of this approach in predicting energy demand, outperforming classic methods in terms of accuracy. The advantage of the proposed approach lies in its ability to present complex time series in a clear nested structure.
Article
Medicine, General & Internal
Ioannis D. Apostolopoulos, Nikolaos D. Papathanasiou, Nikolaos Papandrianos, Elpiniki Papageorgiou, George S. Panayiotakis
Summary: This study evaluates the efficiency of deep learning methods in revealing and suggesting potential image biomarkers. The research concludes that deep learning can reveal potential biomarkers in certain cases, especially when trained in domains where low-level features are not sufficient for decision-making.
Article
Computer Science, Artificial Intelligence
K. Haritha, M. Judy, Konstantinos Papageorgiou, Vassilis C. Georgiannis, Elpiniki Papageorgiou
Summary: The features of a dataset are crucial in constructing a machine learning model. By removing irrelevant features, the classification process can be expedited. The proposed distributed fuzzy cognitive map method exhibits high accuracy in feature selection.
Article
Mathematics
Alexandra Bousia, Aspassia Daskalopulu, Elpiniki Papageorgiou
Summary: Network infrastructure sharing and mobile traffic offloading are promising technologies for providing energy and cost effective services in HetNets. This paper proposes a resource sharing and offloading algorithm based on a double auction mechanism, where MNOs and third parties cooperate to buy and sell capacity and switch off base stations during low traffic periods to reduce energy requirements and costs.
Article
Chemistry, Multidisciplinary
Agorastos-Dimitrios Samaras, Serafeim Moustakidis, Ioannis D. Apostolopoulos, Elpiniki Papageorgiou, Nikolaos Papandrianos
Summary: This article introduces an explainable computer-aided diagnosis system that can help medical experts accurately diagnose cardiovascular diseases, relieving the burden on the National Healthcare Service. The study utilizes a dataset of biometric and clinical information from 571 patients to analyze the prediction process and the significance of each input datum. The findings are compared with the medical literature to evaluate the validity of the prediction process.
APPLIED SCIENCES-BASEL
(2023)
Article
Chemistry, Multidisciplinary
Anna Feleki, Ioannis D. Apostolopoulos, Serafeim Moustakidis, Elpiniki I. Papageorgiou, Nikolaos Papathanasiou, Dimitrios Apostolopoulos, Nikolaos Papandrianos
Summary: This study introduces a novel, transparent, and explainable model called DeepFCM, designed to diagnose Coronary Artery Disease (CAD) using imaging and clinical data. The model utilizes an inner Convolutional Neural Network (CNN) to classify imaging and combines the predictions with clinical data using an FCM-based classifier to determine the presence of CAD. The model's key advantage lies in its explainability, achieved through highlighting significant regions, disclosing internal weights, and generating meaningful explanations.
APPLIED SCIENCES-BASEL
(2023)
Review
Computer Science, Artificial Intelligence
Ioannis D. Apostolopoulos, Nikolaos I. Papandrianos, Elpiniki I. Papageorgiou, Dimitris J. Apostolopoulos
Summary: This study investigates the use of artificial intelligence methods in thyroid and parathyroid surgeries, finding that AI can aid in the localization and identification of parathyroid abnormalities and provide preoperative diagnostic assistance.
MACHINE LEARNING AND KNOWLEDGE EXTRACTION
(2022)
Article
Management
Sadra Ahmadi, Sajjad Shokouhyar, Mohammad Hossein Shahidzadeh, I. Elpiniki Papageorgiou
Summary: This paper presents a framework that utilizes sentiment analysis algorithms to analyze consumer feedback, helping managers make effective decision strategies in reverse logistics. Applying this framework can minimize waste, cost, and inventory, while maximizing efficiency and customer satisfaction.
INTERNATIONAL JOURNAL OF LOGISTICS-RESEARCH AND APPLICATIONS
(2022)