Article
Biochemical Research Methods
Fahimeh Motamedi, Horacio Perez-Sanchez, Alireza Mehridehnavi, Afshin Fassihi, Fahimeh Ghasemi
Summary: This article discusses two approaches for quantitative structure-activity prediction studies, focusing on identifying appropriate molecular descriptors and predicting the biological activities of designed compounds. The use of LASSO-random forest algorithm is shown to significantly improve output correlation, reduce implementation time and model complexity, while maintaining prediction accuracy.
Article
Automation & Control Systems
Yang Xia, Yan Xu, Nan Zhou
Summary: This article proposes a new data-driven method for diagnosing open-circuit faults in multiple inverters in a microgrid. The method uses feature decomposition and extreme learning machines to identify faulty inverters and utilizes a domain adaptation ELM transfer learning algorithm to train the switch fault classifier.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Engineering, Electrical & Electronic
Haneen Bawayan, Mohamed Younis
Summary: This paper proposes an adaptive protection scheme for inverter-based distributed generation (IBDG) in islanded mode. It considers renewable energy sources (RES), Battery Energy Storage System (BESS), and hybrid units in the microgrid (MG). The scheme ensures reliable output power through internal cooperation between RES and BESS and fault current path determination based on power flow analysis. Simulation results on the IEEE 33 bus system validate the effectiveness and superiority of the proposed scheme over competing approaches.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2023)
Article
Business
Hemanta Kumar Bhuyan, Vinayakumar Ravi
Summary: Feature selection is crucial in data mining applications, and most existing methods are insufficient for classification. This article proposes an optimization model based on the Lagrangian multiplier to find and analyze new classes based on subfeature data, and the experiments show its effectiveness.
IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT
(2023)
Article
Computer Science, Artificial Intelligence
Marvin Jimenez, Jose Aguilar, Julin Monsalve-Pulido, Edwin Montoya
Summary: This paper discusses the importance of extracting the correct features from audio and introduces an automated method for analyzing and selecting the best features. The authors developed a hybrid scheme based on different principles for extracting and selecting audio descriptors, and tested it in various audio contexts.
INTERNATIONAL JOURNAL OF MULTIMEDIA INFORMATION RETRIEVAL
(2021)
Article
Construction & Building Technology
Liangdong Ma, Yangyang Huang, Tianyi Zhao
Summary: This paper proposes a synchronous prediction method for predicting building energy consumption, which improves the accuracy of prediction results through feature mining and integrated model for outlier recognition and correction. The method can be widely applied in building energy management systems.
ENERGY AND BUILDINGS
(2022)
Article
Management
Yishi Zhang, Ruilin Zhu, Zhijun Chen, Jie Gao, De Xia
Summary: Feature selection is crucial in fields leveraging big data, and this paper discusses information theoretic methods for feature selection. Two simple lower bounds for feature redundancy and complementarity are introduced, leading to the proposal of a new feature selection method that shows promising improvement in real-world datasets compared to existing methods.
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH
(2021)
Article
Biochemical Research Methods
Souvik Seal, Debashis Ghosh
Summary: This study presents a new method for analyzing the interaction of proteins or markers in the tumor microenvironment of cancer patients, finding co-expression to be associated with disease risk and survival, and demonstrating the robustness of the method.
Article
Computer Science, Artificial Intelligence
Misuk Kim
Summary: Real-time processing and analysis of data are crucial in financial markets due to their direct impact on profits. This study introduces a novel data mining framework focusing on interpretability, prediction metrics, and reporting methods, which was applied to financial prediction problems with successful results.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Information Systems
M. A. N. D. Sewwandi, Yuefeng Li, Jinglan Zhang
Summary: This study introduces a novel method, HCluG, to improve granule identification of continuous data by combining hierarchical clustering with neighborhood rough sets, reducing user involvement. Experimental results show that HCluG can reduce the number of features while improving classification performance.
INFORMATION SCIENCES
(2021)
Article
Automation & Control Systems
Zhanjun Huang, Zhanshan Wang, Chonghui Song
Summary: This study proposes a complementary virtual mirror fault diagnosis method for microgrid inverters, which includes steps such as constructing virtual mirrors, cross-comparison processing, and calculating fault degrees to reduce the impact of data loss and false fault features on fault diagnosis. The joint decision of SCF and AF is utilized for fault detection and localization, demonstrating certain robustness and practicality for inverter fault diagnosis.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Engineering, Electrical & Electronic
Liwei Du, Zhihong Xu, Hongda Chen, Duanyu Chen
Summary: This article proposes a feature selection method to optimize the performance of arc fault diagnosis by reducing data redundancy through permutation importance evaluation criterion and feature clustering. The experimental results demonstrate the effectiveness of this method in improving the accuracy of arc fault diagnosis, and its feasibility in real-world applications.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Review
Energy & Fuels
Zhelun Chen, Zheng O'Neill, Jin Wen, Ojas Pradhan, Tao Yang, Xing Lu, Guanjing Lin, Shohei Miyata, Seungjae Lee, Chou Shen, Roberto Chiosa, Marco Savino Piscitelli, Alfonso Capozzoli, Franz Hengel, Alexander Kuehrer, Marco Pritoni, Wei Liu, John Clauss, Yimin Chen, Terry Herr
Summary: This paper reviews and summarizes the literature on data-driven fault detection and diagnostics (FDD) for building HVAC systems, focusing on the process, systems studied, and evaluation metrics. It identifies challenges such as real-building deployment, performance evaluation, scalability, interpretability, cyber security, data privacy, and user experience that data-driven FDD methods still face despite promising performance reported in the literature.
Article
Computer Science, Information Systems
Shazia Baloch, Mannan Saeed Muhammad
Summary: The proposed scheme utilizes Hilbert transform and data mining approach for microgrid protection, achieving high fault detection accuracy and robustness against measurement noise.
Article
Computer Science, Information Systems
Jian Hu, Huan Xie, Yan Lei, Ke Yu
Summary: This paper proposes a light-weight data augmentation method called Lamont to improve the effectiveness of original FL methods and the efficiency of Aeneas. Lamont uses revised LDA to reduce the dimensionality of the coverage matrix and employs SMOTE to generate synthesized failing tests. Experimental results show that Lamont outperforms original FL methods and is more efficient than Aeneas with comparable effectiveness.
INFORMATION AND SOFTWARE TECHNOLOGY
(2023)
Article
Automation & Control Systems
Tao-Hung Chang, Davor Svetinovic
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2020)
Article
Computer Science, Theory & Methods
Muhammad Habib Ur Rehman, Chee Sun Liew, Teh Ying Wah, Muhammad Imran, Khaled Salah, Nidal Nasser, Davor Svetinovic
Summary: The paper introduces an adaptive execution model for mobile data stream mining applications in MECC environments, which is successfully integrated with multiple MDSM applications. Evaluations show that the proposed adaptive execution model outperforms static and dynamic execution models in various metrics.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2021)
Article
Automation & Control Systems
Muhammad Habib ur Rehman, Ahmed Mukhtar Dirir, Khaled Salah, Ernesto Damiani, Davor Svetinovic
Summary: Cross-device federated learning systems use the blockchain-based framework TrustFed to detect model poisoning attacks, enable fair training settings, and maintain device reputation, resulting in better outcomes compared to conventional approaches.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Automation & Control Systems
Abdullah M. Sawas, Wei Lee Woon, V. Ravikumar Pandi, Mostafa F. Shaaban, Hatem H. Zeineldin
Summary: This article proposes a multistage approach to passive islanding detection using a decision tree algorithm. By passing feature sets extracted using different time windows to successive stages of the tree, the method improves the detection speed and accuracy. The algorithm was trained and tested using a database of feature vectors, demonstrating high detection rates within a short period of time.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Artificial Intelligence
Laith Abualigah, Ali Diabat, Davor Svetinovic, Mohamed Abd Elaziz
Summary: This paper presents a newly proposed metaheuristic algorithm, Harris Hawks Optimization (HHO), and its augmented modification called HHMV. By hybridizing with Multi-verse Optimizer, HHMV improves the convergence speed and search mechanisms of conventional HHO in multi-dimensional optimization problems. Experimental results show that HHMV outperforms other methods in terms of exploration and exploitation search mechanisms and convergence speed.
JOURNAL OF INTELLIGENT MANUFACTURING
(2023)
Article
Computer Science, Interdisciplinary Applications
Sabah Suhail, Saif Ur Rehman Malik, Raja Jurdak, Rasheed Hussain, Raimundas Matulevicius, Davor Svetinovic
Summary: This article presents a blockchain-based digital twin framework for securing Cyber-Physical Systems (CPS). By leveraging digital twins for monitoring, simulating, predicting, and optimizing CPS, the security of CPS can be enhanced. However, the reliability of digital twins depends on the integrity and security of data. This article proposes a trusted digital twin framework for securing CPS and demonstrates its feasibility through a proof of concept.
COMPUTERS IN INDUSTRY
(2022)
Article
Engineering, Electrical & Electronic
Ahmad Mohammad Saber, Amr Youssef, Davor Svetinovic, Hatem Zeineldin, Ehab F. El-Saadany
Summary: This paper proposes an Anomaly-Based Scheme (ABS) for detecting false-tripping attacks against Line Current Differential Relays (LCDRs). Using the Isolation Forest algorithm trained on features determined from local current measurements, the ABS can accurately differentiate between real faults and false-tripping attacks, ensuring that trip commands are only issued for non-attacking faults. Simulation results demonstrate that the ABS can effectively detect different categories of cyberattacks, without compromising the accuracy of fault detection and maintaining robustness to changes in the power system's operating point.
IEEE TRANSACTIONS ON SMART GRID
(2022)
Article
Computer Science, Hardware & Architecture
Anton Wahrstatter, Jorao Gomes Jr, Sajjad Khan, Davor Svetinovic
Summary: This study uses unsupervised machine learning to analyze the complete Bitcoin user graph and identify suspicious actors potentially involved in illegal activities. By introducing a novel set of features, the improved clustering method accurately captures illicit activity within the most suspicious clusters.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2023)
Review
Computer Science, Information Systems
Yusra Abdulrahman, Edin Arnautovic, Vladimir Parezanovic, Davor Svetinovic
Summary: This paper extensively investigates the possibilities of integrating blockchain and Artificial Intelligence (AI) technologies in the field of aerospace engineering, with a focus on supply chain management and operational efficiency. The decentralized nature of blockchain has the potential to greatly improve various aspects of aircraft lifecycle management, while AI can revolutionize predictive supply chain models and detection of structural faults. The paper presents a comprehensive overview of the current state, potential applications, challenges, and future research directions in this field, comparing blockchain technology with traditional record management systems in terms of data storage, security, transparency, and traceability advantages. However, legal, regulatory, and technological readiness issues need to be addressed for wider acceptance in the industry. The findings emphasize the importance of targeted research and development to unlock new applications and drive innovation in aerospace engineering. This paper serves as a comprehensive survey for researchers, practitioners, policymakers, and industry stakeholders, illustrating the transformative potential of AI and blockchain in the aerospace sector.
Review
Computer Science, Information Systems
Jorao Gomes Jr, Sajjad Khan, Davor Svetinovic
Summary: Adopting post-quantum security measures can enhance the privacy and security of traditional blockchains, including accelerated transaction verification, clarified mining authorship, and resilience against quantum attacks. However, there is a lack of comprehensive analysis on post-quantum blockchain consensus. This paper aims to bridge this gap by systematically reviewing Post-Quantum Blockchain Consensus (PQBC), providing a comprehensive overview, critical review, and discussing future research directions.
Article
Computer Science, Information Systems
Sajjad Khan, Jorao Gomes, Muhammad Habib ur ur Rehman, Davor Svetinovic
Summary: This paper presents a decentralized federated learning architecture that detects and eliminates participants with adaptive behavior by evaluating the quality of gradients. Experimental results show that the proposed protocol can effectively detect and eliminate participants with adaptive behavior, while centralized federated learning fails to do so.
INTERNET OF THINGS
(2023)
Article
Automation & Control Systems
Ahmad Mohammad Saber, Amr Youssef, Davor Svetinovic, Hatem H. Zeineldin, Ehab F. El-Saadany
Summary: This article proposes a scheme to protect LCDRs from cyberattacks such as direct-false-tripping, fault-masking, and sympathetic-tripping, by utilizing a trained deep neural network to differentiate between authentic and manipulated LCDR measurements. The proposed scheme accurately detects various forms of cyberattacks while maintaining the protective characteristics of LCDRs.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Proceedings Paper
Computer Science, Information Systems
Ahmed Dirir, Khaled Salah, Davor Svetinovic, Raja Jayaraman, Ibrar Yaqoob, Salil S. Kanhere
Summary: This paper proposes a blockchain-based decentralized federated learning system. The system achieves consensus on training configurations through smart contracts, aggregates model updates using a distributed method, and efficiently handles task scheduling, dropouts, and malicious acts.
2022 FOURTH INTERNATIONAL CONFERENCE ON BLOCKCHAIN COMPUTING AND APPLICATIONS (BCCA)
(2022)
Meeting Abstract
Psychiatry
Katarina Skopljak, Josefina Gerlach, Darjan Svetinovic, Ivan Barun, Igor Filipcic
PSYCHIATRIA DANUBINA
(2022)
Meeting Abstract
Psychiatry
Natko Geres, Sandra Zecevic Penic, Zeljko Milovac, Strahimir Sucic, Tomislav Gajsak, Barbara Koret, Josefina Gerlach, Diana Prskalo-Cule, Darjan Svetinovic, Nikolina Tunjic-Vukadinovic, Ivana Orgulan, Vladimir Pozgaj, Andro Kosec, Goran Geber, Goran Ivkic, Sandra Vuk Pisk, Igor Filipcic
PSYCHIATRIA DANUBINA
(2022)