Article
Computer Science, Information Systems
Branislava Cvijetic, Zaharije Radivojevic
Summary: Tabular data is a common format to represent real-world information, and the problem of importing messy files or files with multiple tables is challenging. This paper proposes the STCExtract algorithm to reconstruct table structures and data, achieving high accuracy according to the evaluation results.
Article
Engineering, Chemical
Kirill Mikhaylov, Stelios Rigopoulos, George Papadakis
Summary: The study combines reduced order modeling and system identification to reconstruct the temporal evolution of large-scale vortical structures behind the blades of a Rushton impeller. The results demonstrate that even using the velocity time signal from a single sensor point, the first pair of modes can be reconstructed well. Increasing the number of sensor points improves accuracy and stability, leading to better reconstruction of the second pair of POD modes. The estimator derived at Reynolds number 600 shows robustness when applied to flows at Reynolds numbers 500 and 700.
Article
Business
Feng Wu, Xin Huang, Bin Jiang
Summary: This article proposes a data-driven approach to automatically identify a subset of the original dataset that can cover more themes and content. The approach improves the accuracy of similarity estimation by incorporating external knowledge and attribute interactions, and identifies representative objects using an enhanced density peaks clustering algorithm. Experimental results demonstrate the effectiveness and robustness of the proposed approach.
IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT
(2022)
Article
Computer Science, Artificial Intelligence
Wentao Li, Haoxiang Zhou, Weihua Xu, Xi-Zhao Wang, Witold Pedrycz
Summary: This article proposes a method for dealing with interval ordered data using interval-valued dominance relation. It first introduces new thresholds for interval dominance degree and interval overlap degree, and then constructs the interval-valued dominance relation. Based on this relation, the interval-valued dominance-based rough set approach and its properties are investigated, and feature selection rules and algorithms are provided.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Information Systems
Zixiang Wang, Tongliang Li, Zhoujun Li
Summary: Numerical information is important in various fields and existing methods for extraction often have limitations. This study proposes QuantityIE, a new approach that leverages syntactic features to extract structured representations of numerical information. Experimental results show that QuantityIE outperforms existing methods in terms of accuracy.
Article
Genetics & Heredity
Michael H. Guo, Laurent C. Francioli, Sarah L. Stenton, Julia K. Goodrich, Nicholas A. Watts, Moriel Singer-Berk, Emily Groopman, Philip W. Darnowsky, Matthew Solomonson, Samantha Baxter, Grace Tiao, Benjamin M. Neale, Joel N. Hirschhorn, Heidi L. Rehm, Mark J. Daly, Anne O'Donnell-Luria, Konrad J. Karczewski, Daniel G. MacArthur, Kaitlin E. Samocha
Summary: This study developed a strategy to infer the phase for rare variant pairs within genes using genotypes observed in the Genome Aggregation Database. The approach showed high accuracy in determining phase in both trio data and patients with Mendelian conditions, providing a valuable resource for interpreting rare co-occurring variants in the context of recessive diseases.
Article
Environmental Sciences
Sandhi Wangiyana, Piotr Samczynski, Artur Gromek
Summary: Building footprints are essential for mapping and disaster management, with SAR images presenting challenges in interpretation. Geometric transformations are more effective than pixel transformations in improving object detection, but should be used moderately to prevent unwanted transformations. Error analysis is recommended to understand dataset biases and guide the selection of suitable DA methods for future research.
Article
Computer Science, Artificial Intelligence
Chuan Luo, Sizhao Wang, Tianrui Li, Hongmei Chen, Jiancheng Lv, Zhang Yi
Summary: This article proposes a novel global search method for numerical feature selection, RH-BPSO, based on the hybridization of the rough hypercuboid approach and binary particle swarm optimization (BPSO) algorithm. Parallelization approaches for large-scale datasets are also presented by decomposing and recombining hypercuboid equivalence partition matrix. The experimental results indicate that RH-BPSO outperforms other feature selection algorithms in terms of classification accuracy, the cardinality of the selected feature subset, and execution efficiency. Moreover, the distributed meta-heuristic optimized rough hypercuboid feature selection algorithm, DiRH-BPSO, is significantly faster than its sequential counterpart and can handle large-scale feature selection tasks on distributed-memory multicore clusters.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Brenda Santana, Ricardo Campos, Evelin Amorim, Alipio Jorge, Purificacao Silvano, Sergio Nunes
Summary: This paper presents a thorough survey of narrative extraction from text, emphasizing the importance and complexity of the annotation process and the use of formal narrative representation structures. It also highlights important open issues in the field of narrative extraction that are yet to be explored.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Review
Biochemical Research Methods
Stefano Castellana, Tommaso Biagini, Luca Parca, Francesco Petrizzelli, Salvatore Daniele Bianco, Angelo Luigi Vescovi, Massimo Carella, Tommaso Mazza
Summary: This study found that hundreds of human proteins interact with degenerated DNA sequences, and identifying these motifs and genomic sites is a challenging research goal in modern molecular biology and bioinformatics. Over the past twenty years, there has been an explosion of computational tools for this task, and sixteen of them were evaluated for their ability to identify known motifs in simulated sequence datasets.
BRIEFINGS IN BIOINFORMATICS
(2021)
Review
Biochemical Research Methods
Ju Xiang, Jiashuai Zhang, Yichao Zhao, Fang-Xiang Wu, Min Li
Summary: This article reviews the progress of disease association research in recent years and introduces biomedical data, computational methods, and tools/platforms for evaluating disease associations. It is found that different types of data can provide a comprehensive perspective to help understand complex diseases.
BRIEFINGS IN BIOINFORMATICS
(2022)
Review
Medicine, General & Internal
Ben Gordon, Clara Fennessy, Susheel Varma, Jake Barrett, Enez McCondochie, Trevor Heritage, Oenone Duroe, Richard Jeffery, Vishnu Rajamani, Kieran Earlam, Victor Banda, Neil Sebire
Summary: This study objectively evaluated freely available data profiling software tools for their applicability in healthcare data. Several tools showed high potential and functionality for use with healthcare datasets. In a synthetic dataset of 1000 patients, two tools consistently performed well across multiple tasks including completeness, consistency, uniqueness, validity, accuracy, and distribution metrics.
Article
Mechanics
Saviz Mowlavi, Mattia Serra, Enrico Maiorino, L. Mahadevan
Summary: This study introduces a method for identifying Lagrangian coherent structures in sparse and noisy trajectory datasets, computing hyperbolic and elliptic LCSs, and demonstrating their accuracy. By deploying these methods on various fluid and experimental datasets, their practicality is showcased.
JOURNAL OF FLUID MECHANICS
(2022)
Article
Biotechnology & Applied Microbiology
Ramyar Molania, Momeneh Foroutan, Johann A. Gagnon-Bartsch, Luke C. Gandolfo, Aryan Jain, Abhishek Sinha, Gavriel Olshansky, Alexander Dobrovic, Anthony T. Papenfuss, Terence P. Speed
Summary: Accurate identification and removal of unwanted variation in RNA-seq data are crucial for meaningful downstream analyses. Our PRPS strategy with RUV-III normalization effectively addresses this issue and can be applied to integrate and normalize large transcriptomic datasets from multiple sources.
NATURE BIOTECHNOLOGY
(2022)
Article
Engineering, Multidisciplinary
XianJia Chen, Zheng Yuan, Qiang Li, ShouGuang Sun, YuJie Wei
Summary: For complex engineering systems such as high-speed trains, load spectra are crucial for safety design and fault diagnoses. This study demonstrates that machine learning modeling combined with limited experimental data can accurately reproduce the history-dependent load spectra in critical sites. A segmentation and randomization strategy is proposed for processing long-duration historical data, improving the accuracy of the data-driven model for long-term load-time history prediction.
SCIENCE CHINA-TECHNOLOGICAL SCIENCES
(2023)
Review
Pharmacology & Pharmacy
Karlie R. Sharma, Christine M. Colvis, Griffih P. Rodgers, Douglas M. Sheeley
Summary: There are many genes within the druggable genome that have not been studied, and the US National Institutes of Health's program provides resources to explore these genes, with the potential for rapid impact on human health.
DRUG DISCOVERY TODAY
(2024)
Review
Pharmacology & Pharmacy
Mohammad Sameer Khan, B. H. Jaswanth Gowda, Waleed H. Almalki, Tanuja Singh, Amirhossein Sahebkar, Prashant Kesharwani
Summary: Mitochondria-specific functional liposomes hold great potential for cancer therapy. This review discusses the association between mitochondria and tumor formation, as well as the advantages of liposomes in delivering drugs to mitochondria.
DRUG DISCOVERY TODAY
(2024)
Review
Pharmacology & Pharmacy
Choong Yong Ung, Cristina Correia, Hu Li, Christopher M. Adams, Jennifer J. Westendorf, Shizhen Zhu
Summary: With increasing human life expectancy, the global medical burden of chronic diseases is growing. Chronic diseases often involve malfunctioning of multiple organs, and understanding the interorgan crosstalk is crucial to understanding the etiology of chronic diseases. Researchers have proposed the locked-state model (LoSM) and cutting-edge systems biology and artificial intelligence strategies to decipher chronic multiorgan locked states. The findings have important clinical implications for improving treatments for chronic diseases.
DRUG DISCOVERY TODAY
(2024)