A5100330330- Advanced Multi-Objective Optimization Algorithms
- Metaheuristic Optimization Algorithms Research
- Evolutionary Algorithms and Applications
- Scheduling and Optimization Algorithms
- Optimization and Search Problems
- Advanced Neural Network Applications
- Advanced Manufacturing and Logistics Optimization
- Assembly Line Balancing Optimization
- Domain Adaptation and Few-Shot Learning
- Anomaly Detection Techniques and Applications
- Optimal Experimental Design Methods
- RNA Research and Splicing
- Nonlinear Waves and Solitons
- Nonlinear Photonic Systems
- Optimization and Packing Problems
- Advanced Image and Video Retrieval Techniques
- Cloud Computing and Resource Management
- Probabilistic and Robust Engineering Design
- Data Stream Mining Techniques
- Advanced Mathematical Physics Problems
- Human Pose and Action Recognition
- Underwater Acoustics Research
- Time Series Analysis and Forecasting
- Software System Performance and Reliability
- Fault Detection and Control Systems
Huazhong University of Science and Technology
2006-2025
Xiangya Hospital Central South University
2025
Central South University
2025
Taiyuan University of Technology
2024-2025
Jiangsu University of Science and Technology
2017-2025
Huaibei Normal University
2024-2025
Merck (Singapore)
2025
Biological E (India)
2025
Shandong University of Science and Technology
2024
University of Science and Technology of China
2020-2024
Surrogate-assisted evolutionary algorithms (SAEAs) have been developed mainly for solving expensive optimization problems where only a small number of real fitness evaluations are allowed. Most existing SAEAs designed low-dimensional single or multiobjective problems, which not well suited many-objective optimization. This paper proposes surrogate-assisted algorithm that uses an artificial neural network to predict the dominance relationship between candidate solutions and reference instead...
In this paper, we propose a framework to accelerate the computational efficiency of evolutionary algorithms on large-scale multiobjective optimization. The main idea is track Pareto optimal set (PS) directly via problem reformulation. To begin with, algorithm obtains reference directions in decision space and associates them with weight variables for locating PS. Afterwards, original optimization reformulated into low-dimensional single-objective problem. problem, reconstructed by objective...
Offspring generation plays an important role in evolutionary multiobjective optimization. However, generating promising candidate solutions effectively high-dimensional spaces is particularly challenging. To address this issue, we propose adaptive offspring method for large-scale First, a preselection strategy proposed to select balanced parent population, and then these are used construct direction vectors the decision reproducing solutions. Specifically, two kinds of adaptively generate...
Only a small number of function evaluations can be afforded in many real-world multiobjective optimization problems (MOPs) where the are economically/computationally expensive. Such pose great challenges to most existing evolutionary algorithms (EAs), which require large for optimization. Surrogate-assisted EAs (SAEAs) have been employed solve expensive MOPs. Specifically, certain used build computationally cheap surrogate models assisting process without conducting evaluations. The infill...
Recent advancements in deep neural networks have made remarkable leap-forwards dense image prediction. However, the issue of feature alignment remains as neglected by most existing approaches for simplicity. Direct pixel addition between upsampled and local features leads to maps with misaligned contexts that, turn, translate mis-classifications prediction, especially on object boundaries. In this paper, we propose a module that learns transformation offsets pixels contextually align...
Recently, increasing works have been proposed to drive evolutionary algorithms using machine-learning models. Usually, the performance of such model-based is highly dependent on training qualities adopted Since it usually requires a certain amount data (i.e., candidate solutions generated by algorithms) for model training, deteriorates rapidly with increase problem scales due curse dimensionality. To address this issue, we propose multiobjective algorithm driven generative adversarial...
Diversity preservation is a crucial technique in multiobjective evolutionary algorithms (MOEAs), which aims at evolving the population toward Pareto front (PF) with uniform distribution and good extensity. In spite of many diversity approaches existing MOEAs, most them encounter difficulties tackling complex PFs. This article gives detail introduction to approaches, as well revelation limitations them. To address this proposes multistage MOEA for better performance. The proposed divides...
Ratio error (RE) estimation of the voltage transformers (VTs) plays an important role in modern power delivery systems. Existing RE methods mainly focus on periodical calibration but ignore time-varying property. Consequently, it is difficult to efficiently estimate state VTs real time. To address this issue, we formulate a (TREE) problem into large-scale multiobjective optimization problem, where multiple objectives and inequality constraints are formulated by statistical physical rules...
Effective access to obtain the complex flow fields around an airfoil is crucial in improving quality of supercritical wings. In this study, a systematic method based on generative deep learning developed extract features for depicting and predict steady airfoils. To begin with, variational autoencoder (VAE) network designed representative fields. Specifically, principal component analysis technique adopted realize feature reduction, aiming optimal dimension VAE. Afterward, extracted are...
Abstract Many real-world optimization applications have more than one objective, which are modeled as multiobjective problems. Generally, those complex objective functions approximated by expensive simulations rather cheap analytic functions, been formulated data-driven The high computational costs of problems pose great challenges to existing evolutionary algorithms. Unfortunately, there not any benchmark reflecting yet. Therefore, we carefully select seven from applications, aiming promote...
Abstract Various works have been proposed to solve expensive multiobjective optimization problems (EMOPs) using surrogate-assisted evolutionary algorithms (SAEAs) in recent decades. However, most existing methods focus on EMOPs with less than 30 decision variables, since a large number of training samples are required build an accurate surrogate model for high-dimensional EMOPs, which is unrealistic optimization. To address this issue, we propose SAEA adaptive dropout mechanism....
Background: Acute lung injury (ALI) is characterized by dysfunction of the alveolar epithelial membrane caused acute inflammation and tissue injury. Qingwenzhike (QWZK) prescription has been demonstrated to be effective against respiratory viral infections in clinical practices, including coronavirus disease 2019 (COVID-19) infection. So far, chemical compositions, protective effects on ALI, possible anti-inflammatory mechanisms remain unknown. Methods: In this study, compositions QWZK were...
Abstract Background Bone or brain metastases may develop in 20–40% of individuals with late‐stage non‐small‐cell lung cancer (NSCLC), resulting a median overall survival only 4–6 months. However, the primary tissue's distinctions between bone, and intrapulmonary NSCLC at single‐cell level have not been underexplored. Methods We conducted comprehensive analysis 14 tissue biopsy samples obtained from treatment‐naïve advanced patients bone ( n = 4), 6) 4) metastasis using sequencing originating...
An AU-rich element (ARE) consisting of repeated canonical AUUUA motifs confers rapid degradation to many cytokine mRNAs when present in the 3' untranslated region. Destabilization with AREs (ARE-mRNAs) is consistent interaction ARE-binding proteins such as tristetraprolin and four AUF1 isoforms. However, association AUF1-mRNA decreased ARE-mRNA stability correlative has not been directly tested. We therefore determined whether overexpression isoforms promotes destabilization are limiting...
Recent studies show that disk-based graph computation on just a single PC can be as highly competitive cluster-based computing systems large-scale problems. Inspired by this remarkable progress, we develop VENUS, system which is able to handle billion-scale problems efficiently commodity PC. VENUS adopts novel architecture features vertex-centric "streamlined" processing - the sequentially loaded and update functions are executed in parallel fly. deliberately avoids loading batch edge data...
The evolutionary algorithms (EAs) are a family of nature-inspired widely used for solving complex optimization problems. Since the operators (e.g. crossover, mutation, selection) in most traditional EAs developed on basis fixed heuristic rules or strategies, they unable to learn structures properties problems be optimized. To equip with learning abilities, recently, various model-based (MBEAs) have been proposed. This survey briefly reviews some representative MBEAs by considering three...
In evolutionary multiobjective optimization, the Pareto front (PF) is approximated by using a set of representative candidate solutions with good convergence and diversity. However, most existing algorithms (MOEAs) have general difficulty in approximation PFs complicated geometries. To address this issue, we propose generic modeling method for where shape nondominated estimated training generalized simplex model. On basis front, further develop an MOEA, both mating selection environmental...
Evolutionary algorithms have been used to solve a variety of many-objective optimization problems, where these problems contain more than three conflicting objectives. Most existing evolutionary shown perform well on with regular Pareto optima l fronts, their performance, however, will often considerably deteriorate those whose optimal fronts are irregular, e.g., discontinuous, degenerated and convex. To address this issue, in paper, we propose region division based algorithm, termed RdEA,...
Given a time series S = ((x <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</inf> , y ), (x xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> …) and prescribed error bound ε, the piecewise linear approximation (PLA) problem with max-error guarantees is to construct function f such that |f(x xmlns:xlink="http://www.w3.org/1999/xlink">i</inf> )-y | ≤ ε for all i. In addition, we would like have an online algorithm takes as records arrive in...