Article
Mathematics
Upasana Lakhina, Nasreen Badruddin, Irraivan Elamvazuthi, Ajay Jangra, Truong Hoang Bao Huy, Josep M. M. Guerrero
Summary: This paper proposes an enhanced multi-objective multi-verse optimizer algorithm (MOMVO) for stochastic generation power optimization in a renewable energy-based islanded microgrid framework. The algorithm schedules power among different generation sources to minimize generation costs and power losses in the microgrid. Simulation results demonstrate that MOMVO outperforms other metaheuristic algorithms for multi-objective optimization.
Article
Computer Science, Artificial Intelligence
Pei Liang, Yaping Fu, Kaizhou Gao, Hao Sun
Summary: The application of big data has been widely studied in various fields, with the disassembly process in remanufacturing systems being a critical step that requires the use of stochastic programming methods and efficient intelligent optimization algorithms due to the uncertainty and complexity of data.
COMPLEX & INTELLIGENT SYSTEMS
(2022)
Article
Engineering, Aerospace
Laura Medioni, Yvan Gary, Myrtille Monclin, Come Oosterhof, Gaetan Pierre, Tom Semblanet, Perrine Comte, Kevin Nocentini
Summary: The increasing amount of debris in Low Earth Orbit poses challenges to the sustainability of the space environment. The current recommendations for deorbiting satellites are insufficient, and active debris removal missions are still in their early stages due to high costs. One approach to reduce costs is to remove multiple pieces of debris per mission, which requires optimizing the number of debris removed while minimizing mission time and propellant usage.
ADVANCES IN SPACE RESEARCH
(2023)
Article
Computer Science, Artificial Intelligence
Xiao-Bing Hu, Sheng-Hao Gu, Chi Zhang, Gong-Peng Zhang, Ming-Kong Zhang, Mark S. Leeson
Summary: This paper proposes a novel nature-inspired method, the ripple-spreading algorithm (RSA), for solving multi-objective path optimization problems. The method is able to calculate the complete Pareto front in just a single run, and can be extended to calculate all Pareto optimal paths in dynamical networks. The potential applications of this method in real-world problems are significant.
SWARM AND EVOLUTIONARY COMPUTATION
(2021)
Article
Mathematics
Mohammad H. H. Nadimi-Shahraki, Hoda Zamani, Ali Fatahi, Seyedali Mirjalili
Summary: Moth-flame optimization (MFO) is a simple yet widely used problem solver for different optimization problems. However, MFO and its variants suffer from poor population diversity, resulting in premature convergence and lower solution quality. To address this issue, an enhanced algorithm called MFO-SFR was developed, which utilizes an effective stagnation finding and replacing (SFR) strategy to maintain population diversity during the optimization process. Extensive evaluations on benchmark functions and comparison with competitors demonstrated that the proposed MFO-SFR algorithm outperformed MFO variants and state-of-the-art metaheuristic algorithms in solving complex global optimization problems, with an effectiveness of 91.38%.
Article
Automation & Control Systems
Yi'an Wang, Kun Li, Ying Han, Xinxin Yan
Summary: This paper proposes a method for real-time tracking of dynamic intrusion targets, including target path prediction and multi-UAV path optimization. A trajectory prediction method is proposed to deal with the uncertainty of the target trajectory. A hybrid algorithm is proposed to improve the optimization accuracy of the tracking trajectory, and its effectiveness is validated through simulation experiments.
Article
Chemistry, Analytical
Maja Rosic, Milos Sedak, Mirjana Simic, Predrag Pejovic
Summary: This paper proposes a chaos-enhanced adaptive hybrid butterfly particle swarm optimization algorithm to estimate the position of a passive target. The algorithm combines adaptive strategy and chaos theory to improve performance, and converts the problem into a convex one using semidefinite programming.
Article
Mathematics, Applied
Maksim Dolgopolik
Summary: Research was conducted on finding the distance between ellipsoids in convex and nonconvex cases, proposing ADMM algorithm, heuristic rules, and restarting procedures, which were numerically verified to be effective.
APPLIED MATHEMATICS AND COMPUTATION
(2021)
Article
Environmental Sciences
Mosaad Khadr, Andreas Schlenkhoff
Summary: The study introduces an optimization-simulation framework using implicit stochastic optimization, genetic algorithms, and recurrent neural networks to improve reservoir management. Results from the application at the Bigge reservoir in Germany show promising effectiveness of the GA-ISO-RNN model in simulating and predicting optimal reservoir release.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2021)
Article
Automation & Control Systems
Abraham P. Vinod, Meeko M. K. Oishi
Summary: This study focuses on the stochastic reachability problem in constrained dynamical systems, providing insights on the geometric properties of stochastic reach sets and proposing a scalable algorithm for their computation using convex optimization. The efficacy and scalability of the approach is demonstrated through numerical examples, showing superior performance compared to existing software tools for linear systems verification.
Article
Computer Science, Software Engineering
Damek Davis
Summary: This paper discusses the importance and methods of minimizing finite sums of smooth and strongly convex functions in machine learning. The authors also investigate the acceleration effect of variance reduction on fixed point and root-finding problems involving sums of nonlinear operators.
MATHEMATICAL PROGRAMMING
(2023)
Article
Thermodynamics
Wenqiang Yang, Xinxin Zhu, Qinge Xiao, Zhile Yang
Summary: This paper proposes an improved version of the multi-objective marine predator algorithm (IMOMPA) for solving the optimization of multi-objective dynamic economic-grid fluctuation dispatch (MODEGD). The IMOMPA algorithm improves population diversity, convergence speed, and global search ability. Numerical experiments on benchmark functions and generation units demonstrate the superiority of the IMOMPA algorithm, and plug-in electric vehicles (PEVs) connected to the grid (V2G) can help mitigate grid fluctuations.
Article
Physics, Applied
Wan Chen, Jiahui Fu, Qun Wu
Summary: This article introduces a technique for increasing the directivity of a superdirective antenna by using a complex excitation source, specifically choosing the Huygens source for its zero-back-lobe characteristic. By applying this method, the directivity of the multi-layered cylinder was increased by 20% compared to previous designs, resulting in an extremely small back-lobe. Unlike other proposed designs focusing on material optimizations, this approach prioritizes the form of the excitation, providing a novel method for improving the performance of superdirective antennas.
JOURNAL OF PHYSICS D-APPLIED PHYSICS
(2021)
Article
Computer Science, Interdisciplinary Applications
Rong Zheng, Abdelazim G. Hussien, Raneem Qaddoura, Heming Jia, Laith Abualigah, Shuang Wang, Abeer Saber
Summary: The African vultures optimization algorithm (AVOA) is a metaheuristic inspired by the African vultures' behaviors. However, it suffers from slow convergence rate and local optimal stagnation. In this study, an enhanced version called EAVOA is introduced, using techniques such as representative vulture selection strategy, rotating flight strategy, and selecting accumulation mechanism. EAVOA outperforms other methods in terms of numerical results and convergence curves, and shows practical applicability in engineering design optimization problems and classification tasks.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Sarah E. Shukri, Rizik Al-Sayyed, Amjad Hudaib, Seyedali Mirjalili
Summary: Cloud computing is a popular technology that enables users to remotely access computing resources in a pay-as-you-go model. Task scheduling is a primary challenge in cloud computing environments, with many meta-heuristic algorithms like MVO and PSO being used. The Enhanced Multi-Verse Optimizer (EMVO) proposed in this paper outperforms both MVO and PSO algorithms in terms of minimizing makespan time and increasing resource utilization in cloud environments.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Biochemical Research Methods
Omar Maddouri, Xiaoning Qian, Byung-Jun Yoon
Summary: GCNCC is a graph convolutional network-based approach that can identify highly effective and robust network-based disease markers. By integrating gene expression data with protein interaction data, GCNCC learns reproducible markers with consistently accurate prediction performance across different diseases and platforms.
Article
Multidisciplinary Sciences
Woo Seok Kim, M. Ibrahim Khot, Hyun-Myung Woo, Sungcheol Hong, Dong-Hyun Baek, Thomas Maisey, Brandon Daniels, P. Louise Coletta, Byung-Jun Yoon, David G. Jayne, Sung Il Park
Summary: The authors report an AI-enabled, implantable, multichannel wireless telemetry for photodynamic therapy that enables uniform delivery of multi-wavelength light to tumors. They overcome the limitations of traditional wireless technologies through a thermal/light simulation platform, providing guidelines for effective photodynamic therapy regimen design.
NATURE COMMUNICATIONS
(2022)
Article
Multidisciplinary Sciences
Omar Maddouri, Xiaoning Qian, Francis J. Alexander, Edward R. Dougherty, Byung-Jun Yoon
Summary: This Data in Brief article presents a synthetic dataset for binary classification in the context of Bayesian transfer learning. The dataset can be used for designing and evaluating transfer learning-based classifiers. It considers various combinations of classification settings and simulates a diverse set of feature-label distributions with different learning complexity. The provided dataset is valuable for designing and benchmarking transfer learning schemes for binary classification and estimating classification error.
Article
Materials Science, Multidisciplinary
Yong-Sik Ahn, Byung-Jun Yoon
Summary: The effects of aging heat treatment on localized corrosion behavior in lean duplex stainless steel with 0-2% Mo were investigated. It was found that the precipitation of Cr2N decreased with increasing Mo content and aging time. The pitting potential (E-pit) and critical pitting temperature values decreased drastically with aging time, which can be attributed to the depletion of Cr and N components related to the precipitation of Cr2N at the alpha/gamma-phase boundary.
JOURNAL OF MATERIALS ENGINEERING AND PERFORMANCE
(2023)
Correction
Materials Science, Multidisciplinary
Byung-Jun Yoon, Yong-Sik Ahn
JOURNAL OF MATERIALS ENGINEERING AND PERFORMANCE
(2023)
Article
Automation & Control Systems
Qihua Chen, Xuejin Chen, Hyun-Myung Woo, Byung-Jun Yoon
Summary: This study proposes a scheme to reduce the computational cost of objective-UQ via MOCU based on a data-driven approach. It incorporates a neural message-passing model for surrogate modeling and introduces a novel axiomatic constraint loss that penalizes an increase in estimated system uncertainty. Results show that the proposed approach can accelerate MOCU-based optimal experimental design by four to five orders of magnitude without any visible performance loss.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Biochemical Research Methods
Puhua Niu, Maria J. Soto, Shuai Huang, Byung-Jun Yoon, Edward R. Dougherty, Francis J. Alexander, Ian Blaby, Xiaoning Qian
Summary: TRIMER is a genome-scale modeling pipeline for metabolic engineering applications. It integrates metabolic reactions with a transcription factor-gene regulatory network (TRN) using a Bayesian network (BN) to effectively predict regulated metabolic reactions. This article focuses on sensitivity analysis of metabolic flux prediction for uncertainty quantification of TRN modeling in TRIMER. A computational strategy is proposed to construct the uncertainty class of TRN models, which guides optimal experimental design (OED) to enhance TRN modeling and achieve specific metabolic engineering objectives.
JOURNAL OF COMPUTATIONAL BIOLOGY
(2023)
Proceedings Paper
Acoustics
Mingzhou Fan, Byung-Jun Yoon, Francis J. Alexander, Edward R. Dougherty, Xiaoning Qian
Summary: Accurate detection of infected individuals is crucial in stopping pandemics. This work explores a noisy adaptive group testing design to improve efficiency by selecting optimal groups based on utility functions.
2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP)
(2022)
Article
Computer Science, Artificial Intelligence
Paul Chong, Byung-Jun Yoon, Debbie Lai, Michael Carlson, Jarone Lee, Shuhan He
Summary: Epidemiological models can predict viral spread and health outcomes, but they may overestimate the effectiveness of non-pharmaceutical interventions. Multiple independently developed models can mitigate inaccuracies and incorrect assumptions.
Article
Computer Science, Artificial Intelligence
Omar Maddouri, Xiaoning Qian, Francis J. Alexander, Edward R. Dougherty, Byung-Jun Yoon
Summary: Classification is a crucial task in building intelligent systems, and our proposed Bayesian minimum mean-square error estimator allows accurate evaluation of classifier error in small-sample settings.
Article
Engineering, Electrical & Electronic
Hyun-Myung Woo, Youngjoon Hong, Bongsuk Kwon, Byung-Jun Yoon
Summary: Recent advancements in objective-based uncertainty quantification have shown the importance of goal-driven approaches in quantifying model uncertainty. A key concept is the mean objective cost of uncertainty (MOCU), which is effective for assessing the impact of uncertainty on operational goals. This paper introduces a novel machine learning scheme to accelerate MOCU computation and improve MOCU-based experimental design, with results demonstrating a significant speed improvement without compromising performance.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Ziyu Xiang, Mingzhou Fan, Guillermo Vazquez Tovar, William Trehem, Byung-Jun Yoon, Xiaofeng Qian, Raymundo Arroyave, Xiaoning Qian
Summary: Automatic Feature Engineering (AFE) is aimed at extracting useful knowledge for interpretable predictions, with a focus on materials science applications. Researchers develop and evaluate new AFE strategies using deep Q-networks to explore feature generation trees to address the issue of constructing and searching the entire feature space.
THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE
(2021)
Article
Biochemical Research Methods
Hyun-Myung Woo, Byung-Jun Yoon
Summary: MONACO is a novel and versatile network alignment algorithm that achieves highly accurate pairwise and multiple network alignments through iterative optimal matching of 'local' neighborhoods around focal nodes. Extensive performance assessments on real and synthetic networks, where the ground truth is known, show that MONACO consistently outperforms all other state-of-the-art network alignment algorithms in terms of accuracy, coherence, and topological quality of the aligned network regions. Despite the significantly enhanced alignment accuracy, MONACO remains computationally efficient and scales well with increasing size and number of networks.
Article
Computer Science, Information Systems
Byung-Jun Yoon, Xiaoning Qian, Edward R. Dougherty
Summary: This paper introduces the concept of multi-objective MOCU for quantifying uncertainty in complex systems. Several illustrative examples demonstrate the concept and strengths of this approach.
Article
Computer Science, Information Systems
Youngjoon Hong, Bongsuk Kwon, Byung-Jun Yoon
Summary: This paper discusses the optimal experimental design problem for an uncertain system described by coupled ODEs, aiming to reduce model uncertainty within a limited experimental budget through the development of an OED strategy based on MOCU. The main objective is to identify the optimal experiment that maximally reduces the uncertainty cost, demonstrating the importance of quantifying potential experiments' operational impact in designing optimal experiments.