Article
Chemistry, Multidisciplinary
K. K. Deepika, P. Srinivasa Varma, Ch Rami Reddy, O. Chandra Sekhar, Mohammad Alsharef, Yasser Alharbi, Basem Alamri
Summary: This paper presents a combined approach of PCA-ELM for boiler output forecasting in a thermal power plant. The dimensionality of the input dataset is reduced using PCA, and ELM with different activation functions and neuron numbers is designed. The performance of the developed approach is evaluated in terms of various error metrics. The results demonstrate that the PCA-ELM approach performs well in reducing errors and time.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Hardware & Architecture
Hualong Yu, Houjuan Xie, Xibei Yang, Haitao Zou, Shang Gao
Summary: This paper investigates the online learning problem with dynamically increased classes and extends the OS-ELM to address this issue. Two different scenarios of increased classes are considered and solutions are proposed using alternative output nodes and a hierarchical structure. Simple experiments are conducted to demonstrate the effectiveness and feasibility of the proposed models.
COMPUTERS & ELECTRICAL ENGINEERING
(2021)
Article
Ergonomics
Raj Bridgelall, Denver D. Tolliver
Summary: Railroads play a critical role in a nation's economic health, but accidents result in significant financial losses annually. Derailments are the leading cause of accidents in the U.S. railroad industry, and key factors distinguishing derailments from other accident types include track class, type of movement authority, excess speed, and territory signalization. Machine learning algorithms and feature scoring methods can effectively predict accident types and help railroad companies manage risks based on their unique operating environments.
ACCIDENT ANALYSIS AND PREVENTION
(2021)
Article
Automation & Control Systems
Ho Pham Huy Anh, Cao Van Kien
Summary: This paper introduces a novel advanced single-hidden layer neural extreme learning machine model with nonlinear parameters (ASHLN-ELM), in which hidden and output weighting values are updated simultaneously using adaptively robust rules based on Lyapunov stability principle. The proposed approach guarantees fast convergence and limits state-estimation residual errors to zero without the need for knowledge related to desired weighting values or required approximating error. The efficiency and robustness of the method are demonstrated using typical uncertain hyper-chaotic benchmark systems.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2021)
Article
Computer Science, Information Systems
Zhijie Li, Ningde Jin, Xin Wang, Jidong Wei
Summary: An extreme learning machine-based tone reservation scheme is proposed to reduce the PAPR of OFDM system, achieving comparable performance with low complexity compared to other Neural Network-based algorithms. The algorithm shows advantages of fast learning capability and short training length in simulation results.
IEEE WIRELESS COMMUNICATIONS LETTERS
(2021)
Article
Computer Science, Information Systems
M. Rezwanul Mahmood, Mohammad Abdul Matin, Panagiotis Sarigiannidis, Sotirios K. K. Goudos, George K. K. Karagiannidis
Summary: This paper investigates the application of residual compensation-based ELM (RC-ELM) in designing a receiver for MIMO-NOMA aided IoT systems. By analyzing the BER and EVM performances, the appropriate number of compensation layers for training error minimization is determined. The results show improved BER and EVM performances with the aid of RC-ELM compared to other receivers.
Article
Computer Science, Information Systems
Yuao Zhang, Qingbiao Wu, Jueliang Hu
Summary: In this study, an adaptive regularized extreme learning machine (A-RELM) is proposed, which uses a function instead of a regularization factor to achieve better regularization. Experimental results show the advantages of the algorithm in some benchmark tests.
Article
Engineering, Electrical & Electronic
Meng Sun, Yunjia Wang, Shenglei Xu, Hongchao Yang, Kewei Zhang
Summary: Indoor positioning using the geomagnetic field has become popular due to its infrastructure-free nature and ubiquitous magnetic signals in indoor environments. The extreme learning machine model optimized by the enhanced genetic algorithm can achieve meter-level location accuracy and demonstrate good robustness under different testing conditions.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)
Article
Geochemistry & Geophysics
Qiubo Hu, Wenxiang Xu, Xiaobo Liu, Zhihua Cai, Junjie Cai
Summary: The study introduces a novel spectral-spatial information integration method based on the BF with multispatial domain (MBF), which includes PCA processing, using multiple principal components for BF filtering, and ELM for classification. Experimental results demonstrate that the proposed method improves existing filtering techniques and enhances classification accuracy.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Review
Chemistry, Analytical
Mengge Yang, Jiajia Wang, Siyu Quan, Qiqi Xu
Summary: The thyroid gland, as the largest endocrine organ, is critical to patient health. The combination of spectroscopic approaches and machine deep learning shows promise as a low-cost, fast, and precise diagnostic tool for thyroid diseases. Machine deep learning models, evaluated by precision, sensitivity, specificity, ROC curve, and F1-score, have significant potential for disease diagnosis.
ANALYTICAL LETTERS
(2023)
Article
Environmental Sciences
Allan T. Tejada, Victor B. Ella, Rubenito M. Lampayan, Consorcia E. Reano
Summary: This study developed SVM and ELM models for daily ETo estimation in the Philippines and found that these machine learning models provide more accurate estimates compared to empirical models, with similar modeling performance.
Article
Engineering, Electrical & Electronic
Suman Biswas, Gautam Kumar Mahanti, Nilanjan Chattaraj
Summary: This article investigates the application of the extreme learning machine (ELM) based fault diagnosis technique in analog signal conditioning circuits, and demonstrates its high accuracy through experimentation.
CIRCUITS SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Computer Science, Information Systems
Qiang Fang, Wenzhuo Zhang, Xitong Wang
Summary: This paper introduces a regularized extreme learning machine-based inverse reinforcement learning approach, which combines these two methods to improve navigation performance and enhance generalization ability.
Article
Computer Science, Artificial Intelligence
Gizem Atac Kale, Cihan Karakuzu
Summary: This paper presents two novel networks called Improved Multilayer Extreme Learning Machines (IML-ELM). These networks utilize neuron activations during and after training and employ random assignment of connection weights and leveraging of previous layer's output weight matrix for improved modeling performance. Experimental results show that IML-ELM1 and IML-ELM2 outperform the traditional Multilayer Extreme Learning Machine (ML-ELM) in modeling various benchmark dynamic systems, with better performance and shorter computation time.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Zhongyang Wang, Junchang Xin, Zhiqiong Wang, Huizi Gu, Yue Zhao, Wei Qian
Summary: The study introduces a CADx framework based on Hierarchical Extreme Learning Machine, which utilizes data redundancy reduction methods like CSP and BFN to achieve a good experimental evaluation.
COGNITIVE COMPUTATION
(2021)
Article
Computer Science, Interdisciplinary Applications
Abdelazim G. Hussien, Ali Asghar Heidari, Xiaojia Ye, Guoxi Liang, Huiling Chen, Zhifang Pan
Summary: This study proposes an enhanced variant of the whale optimization algorithm (WOA) called VCSWOA, which combines components from other algorithms. The comprehensive testing and comparison with other algorithms demonstrate that VCSWOA outperforms its peers in terms of performance.
ENGINEERING WITH COMPUTERS
(2023)
Article
Computer Science, Artificial Intelligence
Shengsheng Wang, Bilin Wang, Zhe Zhang, Ali Asghar Heidari, Huiling Chen
Summary: This paper proposes a novel OT-based Class-Aware Sample Reweighting (CASR) method to achieve sample-level fine-grained alignment between multi-source and target. Extensive experiments show that CASR presents significant advantages compared with other MSDA methods, and the visualization analysis further demonstrates the effectiveness of each proposed module.
Article
Computer Science, Interdisciplinary Applications
Shuhui Hao, Changcheng Huang, Ali Asghar Heidari, Huiling Chen, Lingzhi Li, Abeer D. Algarni, Hela Elmannai, Suling Xu
Summary: Early detection and treatment of fast-growing skin cancers can significantly extend patients' lives. Dermoscopy is a reliable tool for detecting skin cancer in its early stages, and the effective processing of digital dermoscopy images is crucial for improving diagnosis accuracy. This paper proposes an improved salp swarm algorithm (ILSSA) method that combines iterative mapping and local escaping operator to address the issue of local optima frequently encountered in existing meta-heuristic algorithms. Moreover, the ILSSA-based multi-threshold image segmentation approach successfully segments dermoscopic images of skin cancer using 2D Kapur's entropy as the objective function and non-local means 2D histogram to represent image information. Benchmark function test experiments demonstrate that ILSSA effectively alleviates the local optimal problem compared to other algorithms. The skin cancer dermoscopy image segmentation experiment further proves the superiority and adaptability of the proposed ILSSA-based method over other existing approaches at different thresholds.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2023)
Article
Computer Science, Interdisciplinary Applications
Weifeng Shan, Xinxin He, Haijun Liu, Ali Asghar Heidari, Maofa Wang, Zhennao Cai, Huiling Chen
Summary: Harris hawks optimization (HHO) is a well-established swarm-based method that works based on multiple dynamic features and various exploratory and exploitative traits. To improve its performance, this paper introduces the Cauchy mutation mechanism into HHO algorithm, creating CMHHO. Experimental results validate that CMHHO outperforms other optimization algorithms in terms of optimization performance.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Xinsen Zhou, Wenyong Gui, Ali Asghar Heidari, Zhennao Cai, Guoxi Liang, Huiling Chen
Summary: Continuous ant colony optimization algorithm incorporates a random following strategy to enhance global optimization performance and effectively handle high-dimensional feature selection problems. The algorithm performs competitively with other state-of-the-art algorithms in benchmark tests and outperforms well-known classification methods on high-dimensional datasets.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Mingjing Wang, Ali Asghar Heidari, Huiling Chen
Summary: To address the curse of dimensionality in high-dimensional medical data, the multi-objective evolutionary algorithm IFMMOEAD is proposed. It considers the number of selected features, classification accuracy, and correlation measures of features to reduce dimensionality. The algorithm has been verified on benchmark tests and applied to cancer gene expression data and clinical data, showing promising results in high-dimensional medical machine learning.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Boyang Xu, Ali Asghar Heidari, Zhennao Cai, Huiling Chen
Summary: This study proposes a variant of the colony predation algorithm (CPA) called Covariance Gaussian cuckoo Colony Predation Algorithm (CGCPA), which employs a designed gaussian cuckoo variable dimensional strategy to enhance population diversity and global search ability, and a covariance matrix adaptation evolution strategy to enhance convergence speed and capture the global optimal solution. Experimental results show that CGCPA outperforms state-of-the-art algorithms in terms of convergence speed and accuracy.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Computer Science, Artificial Intelligence
Xinsen Zhou, Yi Chen, Zongda Wu, Ali Asghar Heidari, Huiling Chen, Eatedal Alabdulkreem, Jose Escorcia-Gutierrez, Xianchuan Wang
Summary: The slime mould algorithm (SMA) is an optimization algorithm that mimics the behavior of slime moulds. To overcome its limitations, a local dimensional mutation strategy and an all-dimensional neighborhood search strategy called LASMA were introduced. Experimental results showed that LASMA outperformed other algorithms in terms of solution accuracy, stability, and convergence speed. A binary version of LASMA called bLASMA was also developed, which showed better performance in optimization and feature selection tasks. LASMA provides a robust and effective solution for various optimization problems.
Article
Mathematical & Computational Biology
Xiuzhi Zhao, Lei Liu, Ali Asghar Heidari, Yi Chen, Benedict Jun Ma, Huiling Chen, Shichao Quan
Summary: The novel coronavirus pneumonia (COVID-19) is a respiratory disease that requires effective diagnostic methods such as X-ray imaging-based diagnosis. This paper proposes an enhanced version of ant colony optimization for continuous domains (MGACO) for highly effective pre-processing of COVID-19 pathological images. Additionally, a method called MGACO-MIS based on MGACO is developed, which demonstrates strong adaptability and high-quality segmentation results compared to other methods.
FRONTIERS IN NEUROINFORMATICS
(2023)
Article
Automation & Control Systems
Chunmei Zhang, Dan Xia, Huiling Chen, Guiling Chen
Summary: This paper focuses on the partial topology identification problem of stochastic multi-group models with multiple dispersal (SMGMMD) by adaptive pinning control. By utilizing graph-theoretic approach and adaptive synchronization techniques, some sufficient criteria on the partial topology identification of SMGMMD with time delay are derived. Moreover, the partial topological structures of SMGMMD without time delay and the entire topological structures of SMGMMD are successfully identified. Finally, coupled Lorenz systems with time delay are employed to validate the feasibility and effectiveness of the proposed theoretical results.
INTERNATIONAL JOURNAL OF CONTROL AUTOMATION AND SYSTEMS
(2023)
Article
Energy & Fuels
Xuemeng Weng, Ali Asghar Heidari, Huiling Chen
Summary: Accurate determination of photovoltaic (PV) parameters is crucial for the reliable operation of solar systems, uninterrupted power supply, and efficient energy management. This paper proposes a novel parameter extraction model using the Q-learning-based multistrategy improved shuffled frog leading algorithm (CRNSFLA). The comprehensive test results show that CRNSFLA outperforms existing algorithms in parameter extraction problems, making it an effective tool for solar cell parameter extractions.
INTERNATIONAL JOURNAL OF ENERGY RESEARCH
(2023)
Article
Computer Science, Artificial Intelligence
Ruoxi Deng, Zhao-Min Chen, Huiling Chen, Jie Hu
Summary: In this paper, a novel deep-learning-based method for refining object contours is proposed. By introducing keypoint-focal loss and regularization loss, as well as integrating a Transformer-style hyper module, the proposed method achieves outstanding performance in contour detection tasks.
Article
Pharmacology & Pharmacy
Hang Su, Dong Zhao, Ali Asghar Heidari, Zhennao Cai, Huiling Chen, Jiayin Zhu
Summary: This study thoroughly examined the impacts of polycyclic aromatic hydrocarbons (PAHs) toxicity on rat features by developing a high-performance optimization method and combining it with a feature selection model. The results showed that the features of genes PXR, CAR, CYP2B1/2, and CYP1A1/2 had the most impact on rats.
BASIC & CLINICAL PHARMACOLOGY & TOXICOLOGY
(2023)
Article
Automation & Control Systems
Chunmei Zhang, Huiling Chen, Qin Xu, Yuli Feng, Ran Li
Summary: This article discusses a class of stochastic hybrid delayed coupled systems with multiple weights, and derives several conditions for asymptotic synchronization and topology identification of the systems based on Kirchhoff's Matrix-Tree Theorem and Lyapunov stability theory.
NONLINEAR ANALYSIS-HYBRID SYSTEMS
(2024)
Article
Environmental Sciences
Wei Zhou, Pengjun Wang, Xuehua Zhao, Huiling Chen
Summary: This paper proposes a boosted atomic search optimization algorithm with a new anti-sine-cosine mechanism to estimate the parameters of photovoltaic models. The introduction of the anti-sine-cosine mechanism improves the accuracy and reliability of parameter estimation for solar cells and components.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2023)