News Item
Multidisciplinary Sciences
Ewen Callaway
Summary: Two years later, following DeepMind's game-changing AI victory in predicting protein structures, researchers are further advancing their studies based on the success of AlphaFold.
Article
Biology
Hussain Ahmed Chowdhury, Dhruba Kumar Bhattacharyya, Jugal Kumar Kalita
Summary: scRNA-seq data analysis allows for identification of novel cells, specific characterization of known cells, and study of cell heterogeneity. Most clustering methods are influenced by user input, but UICPC shows excellent performance without requiring user input.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Review
Pharmacology & Pharmacy
Li Rong Wang, Limsoon Wong, Wilson Wen Bin Goh
Summary: Machine learning models are widely used in drug development. However, the presence of data doppelgangers can affect the reliability of evaluation methods. This study demonstrates the prevalence of data doppelgangers in biomedical data and provides recommendations to mitigate the doppelganger effect.
DRUG DISCOVERY TODAY
(2022)
News Item
Multidisciplinary Sciences
Carrie Arnold
Summary: There are lingering questions regarding the ability of AI tools to truly disrupt the pharmaceutical industry.
Article
Genetics & Heredity
Jonathan Rosenski, Sagiv Shifman, Tommy Kaplan
Summary: By analyzing gene essentiality and gene expression data from over 900 cancer cell lines, we developed machine learning algorithms to accurately predict the essentiality levels of nearly 3000 genes. Our model outperforms current state-of-the-art works in terms of prediction accuracy and the number of genes.
BMC MEDICAL GENOMICS
(2023)
Article
Biochemical Research Methods
An-Phi Nguyen, Stefania Vasilaki, Maria Rodriguez Martinez
Summary: Interpretability is crucial for machine learning models in critical scenarios. We propose FLAN, a structurally constrained deep neural network that processes each input feature separately, allowing users to estimate the effect of each feature independently.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemistry & Molecular Biology
Maria B. Rabaglino, Jan Bojsen-Moller Secher, Marc-Andre Sirard, Poul Hyttel, Haja N. Kadarmideen
Summary: By comparing epigenomic and transcriptomic modifications in internal organs of calves produced through in vitro embryo production (IVP) and multiple ovulation and embryo transfer (MOET), it was found that pathways related to the activation of the hypothalamus-pituitary-gonadal (HPG) axis were enriched in IVP calves. Blood epigenomic data could be used to predict methylation levels of internal organs, and potential biomarkers were identified for embryo origin prediction.
Article
Chemistry, Multidisciplinary
Murat Dener, Gokce Ok, Abdullah Orman
Summary: The study suggests using memory data in malware detection and applying deep learning and machine learning approaches in a big data environment. Results show that the Logistic Regression algorithm achieved the most successful malware detection in memory analysis.
APPLIED SCIENCES-BASEL
(2022)
Article
Statistics & Probability
Yang Zhou, Lirong Xue, Zhengyu Shi, Libo Wu, Jianqing Fan
Summary: This article presents a method of predicting housing vitality using machine learning and easily accessible data. By utilizing energy, nightlight, and land-use data, it is possible to predict the temporal and spatial distribution of occupied houses in fine granularity. The introduction of Factor-Augmented Regularized Model and land-use data mitigates the issues of dependence and heterogeneity in predicting variables.
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
(2022)
Article
Biology
Shudong Wang, Yu Zhang, Yulin Zhang, Wenhao Wu, Lan Ye, Yunyin Li, Jionglong Su, Shanchen Pang
Summary: Single-cell RNA sequencing (scRNA-seq) is a successful technique for identifying cellular heterogeneity. In this study, an unsupervised clustering method called scASGC is proposed, which outperforms classical and state-of-the-art methods, and is able to accurately identify cell subpopulations.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Computer Science, Interdisciplinary Applications
Victor Chang, Ziyang Ji, Qianwen Ariel Xu
Summary: This article demonstrates the use of big data analytics techniques to extract valuable information from raw data on the Dianping website, revealing flavor, environment, and service score as crucial factors in restaurant levels. The J48 model performs best among three models, achieving an accuracy of 88.89%.
Review
Biotechnology & Applied Microbiology
Parminder S. Reel, Smarti Reel, Ewan Pearson, Emanuele Trucco, Emily Jefferson
Summary: With the advancement of high-throughput omics technologies, it is crucial for biomedical research to adopt integrative approaches to analyze diverse omics data using machine learning algorithms. This can lead to the discovery of novel biomarkers and improve disease prediction and precision medicine delivery.-Methods in machine learning have enabled researchers to gain a deeper insight into biological systems and provide recommendations for interdisciplinary professionals looking to incorporate machine learning skills in multi-omics studies.
BIOTECHNOLOGY ADVANCES
(2021)
Article
Chemistry, Multidisciplinary
Jon Lundstrom, Emma Korhonen, Frederique Lisacek, Daniel Bojar
Summary: LectinOracle model, combining transformer-based representations for proteins and graph convolutional neural networks for glycans, is able to predict protein-glycan interactions accurately and generalize well to new glycans and lectins. It has various applications in improving lectin classification, accelerating lectin directed evolution, predicting epidemiological outcomes, and analyzing host-microbe interactions.
Article
Information Science & Library Science
Arpan Kumar Kar, Spyros Angelopoulos, H. Raghav Rao
Summary: While data availability and access used to be challenging for information systems research, the growth and ease of access to large datasets and data analysis tools has increased interest in using such resources for publishing. However, these publications often lack strong theoretical contributions and focus only on descriptive analysis of big data. This article addresses the need for theory building with Big Data by providing guidelines for both inductive and deductive approaches, as well as highlighting common pitfalls to avoid.
INTERNATIONAL JOURNAL OF INFORMATION MANAGEMENT
(2023)
Review
Agriculture, Dairy & Animal Science
Giovanni Franzo, Matteo Legnardi, Giulia Faustini, Claudia Maria Tucciarone, Mattia Cecchinato
Summary: In the future, the demand for poultry meat and eggs is predicted to increase with population growth. This expansion brings both opportunities and challenges such as pollution, competition for resources, animal welfare concerns, and infectious diseases. Optimization and increased efficiency are needed in poultry production, and the use of big data offers the opportunity to develop tools to maximize farm profitability and reduce impacts. Sensor technologies and advanced statistical techniques are discussed, as well as the progress in pathogen genome sequencing and analysis.
Article
Computer Science, Artificial Intelligence
Zhongyan Zhang, Lei Wang, Yang Wang, Luping Zhou, Jianjia Zhang, Fang Chen
Summary: This paper proposes a novel dataset-driven unsupervised object discovery framework, which utilizes deep feature representation and weakly-supervised object detection to discover objects in the image dataset. The proposed framework improves the performance of region-based instance image retrieval.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Computer Science, Theory & Methods
Triet H. M. Le, Huaming Chen, M. Ali Babar
Summary: This article highlights the importance of SV assessment and prioritization and provides a taxonomy of past research efforts and best practices. The article also discusses current limitations and proposes potential solutions.
ACM COMPUTING SURVEYS
(2023)
Article
Computer Science, Artificial Intelligence
Jaber Valizadeh, Shadi Boloukifar, Sepehr Soltani, Ehsan Jabalbarezi Hookerd, Farzaneh Fouladi, Anastasia Andreevna Rushchtc, Bo Du, Jun Shen
Summary: The study presents a solution for the challenges in the vaccine supply chain through a robust optimization model, considering various costs and risks in the public vaccination program. Numerical experiments based on the vaccine supply chain in Kermanshah, Iran, show that the proposed model significantly reduces mortality risk, inequality in vaccine distribution, and total cost.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Engineering, Civil
Wei Wei, Jun Shen, Akbar Telikani, Mahdi Fahmideh, Wei Gao
Summary: IoV networks, which utilize sensors deployed on vehicles, are a new generation of IoT networks. The use of Fractional Critical Deleted Graph (FCDG) can enhance network stability and reliability, aiding in the recovery or reconfiguration of lost links.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Environmental Studies
Cheng Zhang, Bo Du, Zuduo Zheng, Jun Shen
Summary: This paper provides a systematic review of studies on the interactions between pedestrians and MMVs in shared spaces. The findings suggest that space sharing is feasible but more efforts are needed on data collection and further investigations on various MMV types and different conditions.
TRANSPORTATION RESEARCH PART D-TRANSPORT AND ENVIRONMENT
(2023)
Article
Engineering, Electrical & Electronic
Wei Wei, Bochao An, Ke Qiao, Jun Shen
Summary: This paper proposes a blockchain based Multi-users Oblivious Data Sharing scheme (MODS) for the digital twin system in the context of Industrial Internet of Things (IIoT). MODS combines trusted hardware and cryptography to achieve a balance between security and efficiency, addressing the security threats posed by centralized data sharing services for sensitive DT data.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
A. Zhiwei Ye, B. Ruihan Li, C. Wen Zhou, D. Mingwei Wang, E. Mengqing Mei, F. Zhe Shu, G. Jun Shen
Summary: This paper proposes two innovative feature selection methods that integrate ant colony optimization (ACO) algorithm and hybrid rice optimization (HRO) to address the issue of redundant or irrelevant features in high-dimensional data analysis.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Hanyue Zhang, Li Li, Jun Shen
Summary: In this paper, a novel model called PGLR is proposed for entity recognition and relation extraction. The model addresses information redundancy by generating a pseudo-graph and employs a label reuse approach to aggregate predicted labels and external knowledge. The reconstructed representations are evaluated using a gating mechanism. Experimental results demonstrate that the proposed model achieves higher accuracy, faster inference speed, and requires fewer parameters.
2023 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, IJCNN
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Shuaishuai Zu, Li Li, Jun Shen
Summary: This paper proposes a model called CAKT that combines contrastive learning with attention networks for interpretable knowledge tracing. By using contrastive learning as the training goal, more representative representations of knowledge states and learning interactions can be obtained. Extensive experiments demonstrate the excellent predictive performance of CAKT and the positive effects of considering the two properties. Additionally, CAKT exhibits high interpretability for captured knowledge states.
2023 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, IJCNN
(2023)
Article
Computer Science, Information Systems
Dong Lu, Yanlong Zhai, Jun Shen, Mahdi Fahmideh, Jianqing Wu, Jude Tchaye-Kondi, Liehuang Zhu
Summary: Edge intelligence combines edge computing and deep learning to bring AI to the network's edge. It has received attention for its low network latency and better privacy preservation. However, the inference of deep neural networks is computationally demanding and results in poor real-time performance, making it challenging for resource-constrained edge devices.
IEEE TRANSACTIONS ON SERVICES COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Wanqi Yang, Like Xin, Lei Wang, Ming Yang, Wenzhu Yan, Yang Gao
Summary: In real applications, unpaired multiview data, where samples between views cannot be matched, is a challenging problem in multiview clustering. Existing methods rely on matched samples, but we propose an iterative unpaired multiview clustering strategy (IUMC) to learn a complete and consistent subspace representation among views. We also design two effective UMC methods based on IUMC, which outperform state-of-the-art methods in terms of clustering performance and applicability in incomplete multiview clustering.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Qingguo Zhou, Ming Lei, Peng Zhi, Rui Zhao, Jun Shen, Binbin Yong
Summary: In this paper, we propose novel approaches to improve the security and anti-attack capability of the RangeNet++ model in the field of autonomous driving. One is to calculate the local geometry based on the range image, which can reflect the surface shape of the point cloud. The other is to obtain a general adversarial sample based on the range image that is related only to the image itself and closer to the real world, and add it into the training set for training. Experimental results show that the proposed approaches can effectively improve the RangeNet++ model's defense ability against adversarial attacks and enhance its robustness.
COMPUTER VISION - ACCV 2022 WORKSHOPS
(2023)
Article
Computer Science, Artificial Intelligence
Wenbin Li, Lei Wang, Xingxing Zhang, Lei Qi, Jing Huo, Yang Gao, Jiebo Luo
Summary: This article investigates defensive few-shot learning, a challenging problem in learning robust few-shot models against adversarial attacks. Existing adversarial defense methods cannot effectively solve this problem due to the inconsistency in sample-level distribution between training and test sets in few-shot learning. To overcome this, the authors propose a general defensive few-shot learning framework that addresses two key questions: how to transfer adversarial defense knowledge across different sample distributions, and how to narrow the distribution gap between clean and adversarial examples in few-shot setting. Experimental results show that the proposed framework effectively enhances robustness of existing few-shot models against adversarial attacks. Code is available at https://github.com/WenbinLee/DefensiveFSL.git.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Like Xin, Wanqi Yang, Lei Wang, Ming Yang
Summary: In this article, the authors investigate the issue of unpaired multi-view clustering and propose an effective method called selective contrastive learning for UMC (scl-UMC) to address the challenging problems of uncertain clustering structure and pairing relationship between views. Extensive experiments demonstrate the efficiency of the proposed method.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Information Systems
Lei Qi, Lei Wang, Yinghuan Shi, Xin Geng
Summary: In this paper, a novel method called MixNorm is proposed for tackling the generalizable multi-source person re-identification task. The method enhances the diversity of features through domain-aware mix-normalization and domain-aware center regularization, thereby boosting the model's generalization capability in unseen domains. Extensive experiments validate the effectiveness of the proposed method and demonstrate its superiority over state-of-the-art methods.
IEEE TRANSACTIONS ON MULTIMEDIA
(2023)