Article
Computer Science, Information Systems
Zichen Zhang, Shifei Ding, Yuting Sun
Summary: This paper introduces a new method called multiple birth support vector regression (MBSVR), which constructs the regressor from multiple hyperplanes obtained by solving small quadratic programming problems, aiming for faster computation and better fitting precision.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Lingyu Li, Zhi-Ping Liu
Summary: In this paper, the authors propose a CNet-SVM method for discovering biomarker genes from high-throughput omics data. This method can maintain the connectivity between genes while selecting features, and has shown good classification and prediction capabilities on simulation datasets and real-world breast cancer datasets. The results demonstrate the effectiveness of CNet-SVM in selecting connected-network-structured features from high-throughput data.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Optics
Rendong Ji, Yue Han, Xiaoyan Wang, Haiyi Bian, Jiangyu Xu, Zhezhen Jiang, Xiaotao Feng
Summary: This study efficiently predicted pesticide residue content by combining fluorescence spectroscopy with support vector machine (SVM), optimizing the SVM algorithm through K-fold cross validation and grid search. The reliability of the results was confirmed by performance evaluation index and running time.
Article
Optics
Yinshan Yu, Mingzhen Shao, Lingjie Jiang, Yongbin Ke, Dandan Wei, Dongyang Zhang, Mingxin Jiang, Yudong Yang
Summary: In this study, support vector machine (SVM) was used for spectral detection of multiple components in the water environment system. The results showed that the proposed method has high accuracy and robustness, making it suitable for quantitative analysis of multiple components.
Article
Computer Science, Information Systems
Guoquan Li, Linxi Yang, Zhiyou Wu, Changzhi Wu
Summary: Proximal support vector machine (PSVM) is a variant of support vector machine (SVM) which aims to generate a pair of non-parallel hyperplanes for classification. Introducing l(0)-norm regularization in PSVM enables simultaneous selection of important features and removal of redundant features for classification. The proposed method utilizes a continuous nonconvex function and difference of convex functions algorithms (DCA) to solve the optimization problem efficiently.
INFORMATION SCIENCES
(2021)
Article
Environmental Sciences
Guangxin Liu, Liguo Wang, Danfeng Liu, Lei Fei, Jinghui Yang
Summary: This article proposes a non-parallel SVM model, which improves the classification effect and generalization performance for hyperspectral images by adding an additional empirical risk minimization term and bias constraint.
Article
Cell Biology
James H. Felce, Lucia Parolini, Erdinc Sezgin, Pablo F. Cespedes, Kseniya Korobchevskaya, Mathew Jones, Yanchun Peng, Tao Dong, Marco Fritzsche, Dirk Aarts, John Frater, Michael L. Dustin
Summary: The study reveals the spatiotemporal organization of GPCRs within the synapse of T cells, focusing on the contribution of CXCR4 in T cell activation. These GPCRs show a similar pattern of organization in the synapse to CXCR4, providing insights into how they contribute to T cell activation.
FRONTIERS IN CELL AND DEVELOPMENTAL BIOLOGY
(2021)
Article
Biochemical Research Methods
Kexin Hu, Qimeng Sun, Ruifen Chen, Tinghao Xu, Yuncheng Li, Lili Chen, Aidong Wang, Hejing Qi, Danni Shao, Huanning Yue, Yaning Wang, Ziqi Tang, Yi Wang, Chunfeng Liu, Haijun Lv, Fen Wang, Huizhong Xu
Summary: Recently, a new fluorescent covalent staining method has been developed for visualizing anatomical structures in cells and tissues. This method utilizes carboxylate and phosphate stains, which provide higher signal intensity compared to existing stains. It allows more accurate identification of nucleoli in cancer cells and reveals a variety of sub-cellular structures when combined with existing amine stains in expansion microscopy samples. Furthermore, this staining method enables imaging of lipid-based structures in cultured cells, expanding the toolset for histological fluorescence visualization.
JOURNAL OF BIOPHOTONICS
(2023)
Article
Computer Science, Artificial Intelligence
Liming Liu, Maoxiang Chu, Rongfen Gong, Li Zhang
Summary: The improved nonparallel support vector machine (INPSVM) proposed in this article inherits the advantages of nonparallel support vector machine (NPSVM) while also offering incomparable benefits over twin support vector machine (TSVM). INPSVM effectively eliminates noise effects and achieves higher classification accuracy for both linear and nonlinear datasets compared to other algorithms. Experimental results demonstrate the superior efficiency, accuracy, and robustness of INPSVM.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Bagesh Kumar, Ayush Sinha, Sourin Chakrabarti, O. P. Vyas
Summary: In this paper, a fast training method for OCSSVM is proposed, which enhances its scalability without compromising precision significantly. Experimental results show that the proposed method achieves the best tradeoff between training time and accuracy, providing similar accuracies to regular OCSSVM and better scalability compared to existing state-of-the-art one-class classifiers.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Chen Ding, Tian-Yi Bao, He-Liang Huang
Summary: The study proposes a quantum-inspired classical algorithm for LS-SVM, utilizing an improved sampling technique for classification. The theoretical analysis indicates that the algorithm can achieve classification with logarithmic runtime for low-rank, low-condition number, and high-dimensional data matrices.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Interdisciplinary Applications
Bikram Kumar, Deepak Gupta
Summary: The paper introduces a novel method ULTBSVM which utilizes Universum data to enhance the classification of healthy and seizure EEG signals, showing promising results in experiments.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Article
Computer Science, Theory & Methods
Umesh Gupta, Deepak Gupta
Summary: This paper presents two efficient variant models to handle noise and outliers, obtaining solutions by solving a system of linear equations and minimizing the impact of noise. The proposed models demonstrate exceptional generalization performance.
FUZZY SETS AND SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Xiaohan Zheng, Li Zhang, Leilei Yan
Summary: This paper proposes a novel sparse discriminant twin support vector machine (SD-TSVM) to improve the discriminant ability and sparsity compared to traditional TSVM. The introduction of twin Fisher regularization terms and the utilization of 1-norm of coefficients and hinge loss contribute to its satisfactory performance.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Hossein Moosaei, M. A. Ganaie, Milan Hladik, M. Tanveer
Summary: Imbalanced datasets are common in real-world problems. Traditional classification algorithms have limitations in handling imbalanced data. To improve classification performance on imbalanced datasets, an improved reduced universum twin support vector machine (IRUTSVM) algorithm is proposed, which introduces new constraints and reduces computational time.
Article
Energy & Fuels
Andreas W. Momber, Torben Moeller, Daniel Langenkaemper, Tim W. Nattkemper, Daniel Bruen
Summary: This paper introduces a Digital Twin concept for condition monitoring and prescriptive maintenance planning for protective coating systems on wind turbine towers. The concept aims to improve efficiency and reduce human effort by generating reference monitoring areas through a Virtual Twin and integrating various parameters and data. It provides a new maintenance method for tower structures of large onshore wind turbines.
Article
Green & Sustainable Science & Technology
Andreas W. Momber, Tim W. Nattkemper, Daniel Langenkaemper, Torben Moeller, Daniel Bruen, Peter Schaumann, Sulaiman Shojai
Summary: This study proposes a method to utilize digital data for monitoring and maintenance planning of surface protection systems for large onshore wind turbines. The method involves segmenting the wind power structure into reference areas and using multi-modal data to assess each area, thus improving the efficiency of the monitoring process.
Correction
Green & Sustainable Science & Technology
Andreas W. Momber, Tim W. Nattkemper, Daniel Langenkamper, Torben Moller, Daniel Bruen, Peter Schaumann, Sulaiman Shojai
Article
Biochemical Research Methods
Konstantinos Tzanakis, Tim W. Nattkemper, Karsten Niehaus, Stefan P. Albaum
Summary: MetHoS is an automated web-based software platform for processing and analyzing large-scale metabolomics data sets using mass spectrometry. It utilizes a big data framework for parallel processing and distributed storage, enabling comprehensive analysis of hundreds or even thousands of experiments.
BMC BIOINFORMATICS
(2022)
Article
Multidisciplinary Sciences
Daniel Langenkaemper, Aksel Alstad Mogstad, Ingrid Myrnes Hansen, Thierry Baussant, Oystein Bergsagel, Ingunn Nilssen, Tone Karin Frost, Tim Wilhelm Nattkemper
Summary: Hyperspectral imaging (HSI) is a promising technology for environmental monitoring, but its high dimensionality and complexity make the analysis challenging. This study investigated the impact of different stressor exposure patterns on the spectrum of cold water coral through laboratory experiments. A new software tool called HypIX was developed to explore the relationships between spectral signatures and experimental parameters in hyperspectral datasets. The results showed that corals exposed to particles had a larger change rate in spectrum, and the responses varied among coral samples, indicating individual tolerance levels. The proposed HypIX workflows can provide reproducible HSI analysis results.
Article
Chemistry, Analytical
Mingkun Tan, Daniel Langenkamper, Tim W. Nattkemper
Summary: This study investigates the effect of data augmentation in the context of taxonomic classification in underwater images, and finds that the performance of established data augmentation methods differs in marine image collections compared to other image collections. New augmentation strategies are proposed and shown to outperform existing algorithms.
Correction
Multidisciplinary Sciences
Timm Schoening, Jennifer M. Durden, Claas Faber, Janine Felden, Karl Heger, Henk-Jan T. Hoving, Rainer Kiko, Kevin Koeser, Christopher Kraemmer, Tom Kwasnitschka, Klas Ove Moeller, David Nakath, Andrea Nass, Tim W. Nattkemper, Autun Purser, Martin Zurowietz
Article
Multidisciplinary Sciences
Timm Schoening, Jennifer M. Durden, Claas Faber, Janine Felden, Karl Heger, Henk-Jan T. Hoving, Rainer Kiko, Kevin Koeser, Christopher Kraemmer, Tom Kwasnitschka, Klas Ove Moeller, David Nakath, Andrea Nass, Tim W. Nattkemper, Autun Purser, Martin Zurowietz
Summary: Due to the lack of universally adopted data standards, the heterogeneity of underwater image data makes comparison and analysis challenging. To enable sustainable image analysis and processing, standardized formats and procedures are needed along with solutions for data reuse in long-term repositories.
Article
Environmental Sciences
Andrea M. Burfeid-Castellanos, Michael Kloster, Sara Beszteri, Ute Postel, Marzena Spyra, Martin Zurowietz, Tim W. Nattkemper, Bank Beszteri
Summary: This study presents a new method of diatom identification and counting using digital virtual slides, comparing it to the traditional light microscopy workflow. The results show that the digital method provides comparable results in terms of species richness, diatom indices, and community composition. The digital workflow not only improves transparency and reusability, but also increases taxonomic precision, making it a promising alternative to traditional diatom analyses.
Article
Construction & Building Technology
Andreas Momber, Daniel Langenkaemper, Torben Moeller, Tim W. Nattkemper
Summary: The paper presents an online platform for annotation and management of visual digital data for condition monitoring and maintenance modelling of marine steel structures' protective coatings. The authors introduce a 10-step procedure which includes data acquisition, transfer to the online platform, annotation and management using special tools and methods, and modelling for condition monitoring and maintenance purposes. Special platform applications are illustrated as examples.
Article
Engineering, Marine
Andreas W. Momber, Daniel Langenka, Torben Moller, Tim W. Nattkemper
Summary: The application of protective coating systems is a major measure against corrosion of marine steel structures. Digital visual inspection tools are becoming standard for monitoring the condition of coatings on large and complex structures. However, the challenge lies in evaluating large amounts of visual data in a short time. This study introduces an online exploration and annotation tool for coating condition monitoring data and its integration into a prescriptive coating maintenance model.
Article
Multidisciplinary Sciences
Michael W. Kloster, Andrea M. Burfeid-Castellanos, Daniel Langenkaemper, Tim W. Nattkemper, Bank Beszteri
Summary: In this study, deep learning-based segmentation methods were used to segment gigapixel-sized, high-resolution scans of diatom slides with a realistically cluttered background. Object-based tiling approaches were found to improve pixel-based precision and reduce errors in cropping fragments compared to the standard sliding window tiling approach.
Article
Multidisciplinary Sciences
Robin van Kevelaer, Daniel W. Langenkaemper, Ingunn Nilssen, Pal Buhl-Mortensen, Tim Nattkemper
Summary: Fixed underwater observatories (FUO) equipped with sensors are commonly used for recording time series data for marine habitat monitoring. This paper analyzes time series data from two consecutive monitoring campaigns, successfully applying convolutional neural networks (CNN) for the segmentation and classification of coral and polyp activities. The results show differences and similarities between the two time periods, and a time series prediction experiment using recurrent neural networks (RNN) is conducted for predicting polyp activity. Overall, this paper presents important findings for marine biology research.
Review
Environmental Sciences
Alex David Rogers, Hannah Appiah-Madson, Jeff A. Ardron, Nicholas J. Bax, Punyasloke Bhadury, Angelika Brandt, Pier-Luigi Buttigieg, Olivier De Clerck, Claudia Delgado, Daniel L. Distel, Adrian Glover, Judith Gobin, Maila Guilhon, Shannon Hampton, Harriet Harden-Davies, Paul Hebert, Lisa Hynes, Miranda Lowe, Sandy MacIntyre, Hawis Madduppa, Ana Carolina de Azevedo Mazzuco, Anna McCallum, Chris McOwen, Tim Wilhelm Nattkemper, Mika Odido, Tim O'Hara, Karen Osborn, Angelique Pouponneau, Pieter Provoost, Muriel Rabone, Eva Ramirez-Llodra, Lucy Scott, Kerry Jennifer Sink, Daniela Turk, Hiromi Kayama Watanabe, Lauren V. Weatherdon, Thomas Wernberg, Suzanne Williams, Lucy Woodall, Dawn J. Wright, Daniela Zeppilli, Oliver Steeds
Summary: Ocean Census is a strategic science mission aimed at accelerating the discovery and description of marine species. It addresses the knowledge gap of marine biodiversity and the need to protect marine life and ecosystems.
FRONTIERS IN MARINE SCIENCE
(2023)
Article
Multidisciplinary Sciences
Linda C. Muskat, Yannic Kerkhoff, Pascal Humbert, Tim W. Nattkemper, Jorgen Eilenberg, Anant Patel
Summary: A new computer-assisted image analysis method has been developed for the rapid, simple, objective, and reproducible quantification of actively discharged fungal spores. This method can be used with conventional laboratory equipment and open-source software without technical expertise. By establishing a linear relationship between gray value and the automatically counted number of conidia, the gray value can be used as a single parameter for quantification, facilitating quality control of fungal formulations for biological pest control.