Article
Engineering, Biomedical
Ahmed Ebied, Eli Kinney-Lang, Javier Escudero
Summary: This study demonstrates the potential application of higher-order tensor decomposition in proportional myoelectric control and proposes a novel constrained Tucker decomposition (consTD) technique for synergy extraction. The results show that the consTD model outperforms matrix factorisation methods in estimating control signals for wrist movements, providing a more direct and efficient approach for identifying synergies.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2021)
Article
Computer Science, Artificial Intelligence
Yan Jiang, Wei Liang, Jintian Tang, Hongbo Zhou, Kuan-Ching Li, Jean-Luc Gaudiot
Summary: Representation learning techniques have been applied in multimedia content analysis and retrieval. This study presents an efficient multimedia data clustering method that combines sparse coding and manifold regularization to handle global and local information simultaneously, demonstrating its effectiveness and efficiency through comprehensive experiments.
CONNECTION SCIENCE
(2021)
Article
Business
Lifeng Zhang, Xiangrui Chao, Qian Qian, Fuying Jing
Summary: This study examines the innovative development of financial inclusion in the banking industry and proposes a method to supplement missing credit information. Empirical analysis supports the effectiveness of this method.
TECHNOLOGICAL FORECASTING AND SOCIAL CHANGE
(2022)
Article
Computer Science, Artificial Intelligence
Ting Xie, Hua Zhang, Ruihua Liu, Hanguang Xiao
Summary: In this work, a clustering method is proposed by improving the SNMF model and its acceleration algorithm. A novel iterative solution is presented for SNMF with implicit sparse constraints, and a low-dimensional feature space is introduced as the result. Experimental results demonstrate that the improved algorithm performs well in data clustering.
PATTERN RECOGNITION LETTERS
(2022)
Article
Computer Science, Information Systems
Chengliang Liu, Zhihao Wu, Jie Wen, Yong Xu, Chao Huang
Summary: This paper proposes a method called localized sparse incomplete multi-view clustering (LSIMVC) to solve the clustering problem on incomplete multi-view data. LSIMVC learns a sparse and structured consensus latent representation by optimizing a sparse regularized and novel graph embedded multi-view matrix factorization model. It introduces a norm based sparse constraint and a novel local graph embedding term to obtain the individual representations and the consensus representation. It also introduces an adaptive weighted learning scheme to reduce the imbalance factor in incomplete multi-view learning. Experimental results demonstrate that LSIMVC outperforms state-of-the-art IMC approaches.
IEEE TRANSACTIONS ON MULTIMEDIA
(2023)
Article
Automation & Control Systems
Xiangli Li, Xiyan Lu, Xuezhen Fan
Summary: Nonnegative matrix factorization (NMF) is an effective method for high dimensional data analysis, but it cannot utilize label information. To address this, a semi-supervised sparse neighbor constrained co-clustering model (SSCCDS) is proposed. By introducing co-clustering and regularization constraints, SSCCDS overcomes the limitations of traditional NMF and achieves good clustering performance, as demonstrated by experiments on different datasets.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2022)
Article
Computer Science, Artificial Intelligence
Kajal Eybpoosh, Mansoor Rezghi, Abbas Heydari
Summary: This paper presents a method for clustering data points on submanifolds of an unknown manifold. The method maps the intrinsic manifolds to n-spheres using conformal mapping, preserving angles and sparse similarities. The proposed method improves the efficiency of sparse subspace clustering on special manifold data.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2023)
Article
Computer Science, Artificial Intelligence
Xue Li, Sifan Cao, Dan Huang, Ming Zhang, Yiwei Li
Summary: Advances have been made in hyperspectral unmixing with manifold Non-negative Matrix Factorisation methods. However, these methods tend to only consider the preliminary structural information, leading to degraded performance. To overcome this, a novel approach called Global centralised and Structured discriminative Non-negative Matrix Factorisation (GSNMF) was proposed, which effectively represented hyperspectral unmixing by considering both global and local data relationships. Experimental results on synthetic data and real-world datasets demonstrated the superiority of GSNMF compared to state-of-the-art methods in terms of accuracy and consistency in obtaining fractional abundances.
IET COMPUTER VISION
(2023)
Article
Biochemical Research Methods
Rui-Yi Li, Zhiye Wang, Jihong Guan, Shuigeng Zhou
Summary: In this paper, a new clustering method called SPARC is proposed for analyzing scRNA-seq data. The method introduces a novel similarity metric based on sparse representation coefficients to capture the relationships among cells. Experimental results show that SPARC outperforms existing methods and is more effective in mining high quality clusters of scRNA-seq data.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2022)
Article
Geochemistry & Geophysics
Nan Huang, Liang Xiao, Yang Xu, Jocelyn Chanussot
Summary: This study proposes a novel algorithm extending SSC to co-cluster large HSIs to overcome challenges in existing methods. By introducing bipartite graph partitioning in a superpixel and pixel co-clustering framework, the effectiveness of HSI clustering is improved.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Information Systems
Weiyu Guo
Summary: The proposed SDG Deep NMF approach can learn deep features that are sparse and informative, while also sufficiently exploring the local invariance of the data. By performing dual graph regularization in the deep NMF framework, it can respect the intrinsic geometrical structures of the input data in the data and feature spaces.
Article
Geochemistry & Geophysics
Nan Huang, Liang Xiao, Jianjun Liu, Jocelyn Chanussot
Summary: A novel graph convolutional sparse subspace coclustering (GCSSC) model is proposed to address the clustering challenge in large HSIs by integrating affinity matrix learning and spectral coclustering. The use of superpixel-based adaptive dictionary construction and graph convolution techniques improves the accuracy of pixel representation, while a nonnegative orthogonal factorization constraint is introduced to reduce computational and memory consumption. The proposed method is memory and computationally efficient, outperforming state-of-the-art HSI clustering methods in clustering performance.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Artificial Intelligence
Guo Zhong, Chi-Man Pun
Summary: This article proposes a novel matrix factorization framework, called Joint Hypergraph Embedding and Sparse Coding (JHESC), which captures high-order semantic information in data. Experimental results demonstrate that the proposed method consistently outperforms other state-of-the-art matrix factorization methods in data clustering.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Qingjiang Xiao, Shiqiang Du, Kaiwu Zhang, Jinmei Song, Yixuan Huang
Summary: This paper proposes a novel adaptive sparse graph learning method (ASGL) for multi-view spectral clustering, which constructs similarity matrices between views using an adaptive neighbor graph learning method and combines complementary information between views by assigning weights. Experimental results demonstrate the excellent clustering performance of ASGL.
APPLIED INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Zhaoqun Shi, Jinglei Liu
Summary: Clustering has been extensively researched in important fields such as machine learning and data mining. However, traditional NMF-based clustering methods struggle with handling noise and outliers, and the graph regularization approach relies on fixed graph construction. Additionally, there is a lack of sparsity constraints on the basis and coefficient matrices. To address these issues, we propose a joint doubly stochastic matrix regularization and dual sparse coding framework (DSNMF). Experimental results demonstrate the superiority of our method compared to others, as well as its robustness to noise and outliers.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Psychology, Clinical
Tangui Barre, Clemence Ramier, Izza Mounir, Renaud David, Loick Menvielle, Fabienne Marcellin, Patrizia Carrieri, Camelia Protopopescu, Faredj Cherikh
Summary: This study aimed to assess the protective role of dispositional mindfulness on increased tobacco and alcohol use among French hospital workers during the COVID-19 lockdown. The results showed a negative association between dispositional mindfulness and increased tobacco and alcohol use, with mindfulness partially mediating the effect of affect deterioration on tobacco use.
INTERNATIONAL JOURNAL OF MENTAL HEALTH AND ADDICTION
(2023)
Article
Public, Environmental & Occupational Health
Clemence Casanova, Clemence Ramier, Davide Fortin, Patrizia Carrieri, Julien Mancini, Tangui Barre
Summary: This French study found that 10% of adults used CBD, and several user profiles emerged based on demographic and behavioral characteristics. The results suggest the need for clearer European CBD regulations to ensure safe and high-quality products.
Letter
Substance Abuse
Tangui Barre, Helene Venturino, Vincent Di Beo, Patrizia Carrieri, Anne-Laure Pelissier-Alicot
DRUG AND ALCOHOL REVIEW
(2023)
Article
Water Resources
Moustapha Sy, Dimitra Eleftheriadou, Christian Jung, Oliver Lindtner, Spyros Karakitsios, Dimosthenis Sarigiannis, Till Weber, Marike Kolossa-Gehring, Matthias Greiner
Summary: This study aimed to develop an integrative modeling framework to assess the long-term exposure to lead in Belgium, Czech Republic, Germany, and Norway. The framework used estimates of external exposure to predict the concentrations of lead in human blood (PbB) and accounted for past and current exposure events. The modeling approach included a population-based simulation using a two-dimensional Monte Carlo approach to characterize the inter-individual variability and uncertainty in external exposure estimates. The predicted PbB levels were compared with HBM data.
EXPOSURE AND HEALTH
(2023)
Article
Substance Abuse
Tangui Barre, Clemence Ramier, Saskia Antwerpes, Marie Costa, Morgane Bureau, Gwenaelle Maradan, Vincent Di Beo, Christophe Cutarella, Jacques Leloutre, Olivier Riccobono-Soulier, Sophie Hedoire, Elodie Frot, Fabienne Vernier, Stephanie Vassas-Goyard, Sabine Dufort, Camelia Protopopescu, Fabienne Marcellin, Danielle Casanova, Marion Coste, Patrizia Carrieri
Summary: Alcohol use disorder (AUD) is a significant problem in France, but barriers such as stigma and apprehension about total abstinence contribute to a large treatment gap. This study assesses the effectiveness of the Choizitaconso TPE program for people with AUD, finding improvements in internalized stigma.
DRUG AND ALCOHOL REVIEW
(2023)
Editorial Material
Neurosciences
Tangui Barre, Helene Venturino, Vincent Di Beo, Patrizia Carrieri, Anne-Laure Pelissier-Alicot
ENCEPHALE-REVUE DE PSYCHIATRIE CLINIQUE BIOLOGIQUE ET THERAPEUTIQUE
(2023)
Article
Nursing
Tangui Barre, Damien Testa, Melina Santos, Fabienne Marcellin, Perrine Roux, Patrizia Carrieri, Lise Radoszycki, Camelia Protopopescu
Summary: This study aimed to identify factors associated with cannabinoid-based product (CBP) use in multiple sclerosis (MS) patients in France and Spain. The results showed that the prevalence of CBP use was similar in both countries, and the severity of MS was associated with CBP use.
JOURNAL OF CLINICAL NURSING
(2023)
Article
Psychology, Clinical
Martin Bastien, Salim Mezaache, Cecile Donadille, Victor Martin, Laurent Appel, Maela Lebrun, Laelia Briand Madrid, Tangui Barre, Perrine Roux
Summary: Many cannabis users in France reported therapeutic benefits from cannabis consumption even without medical recommendation. A cross-sectional survey was conducted in 2020, collecting data from 4150 daily cannabis users. Approximately 10% of the participants exclusively used cannabis for therapeutic purposes. Exclusive therapeutic users differed from non-exclusive users in terms of age, employment, urban residence, physical and mental health condition, mode of cannabis administration, frequency of use, home cultivation, at-risk alcohol use, and previous-month opiate use. Understanding the profiles of regular cannabis users can inform harm reduction strategies and improve access to care for this population. Further research is needed to clarify the boundaries between therapeutic and recreational use.
JOURNAL OF PSYCHOACTIVE DRUGS
(2023)
Article
Substance Abuse
Tangui Barre, Vincent Di Beo, Perrine Roux, Abbas Mourad, Pierre Verger, Lisa Fressard, Thomas Herault, Jean-Francois Buyck, Francois Beck, Patrizia Carrieri
Summary: Alcohol use is a significant risk factor for premature death and disability. To address this issue, it is important for healthcare systems, particularly general practitioners (GPs), to conduct more frequent and accurate screening for at-risk consumption. This study examined the screening practices of GPs in France and identified factors associated with more frequent screening and the use of validated screening tools. The results showed an improvement in screening practices, but training and perceptions still hinder systematic and accurate screening.
ALCOHOL AND ALCOHOLISM
(2023)
Letter
Gastroenterology & Hepatology
Tangui Barre, David Zucman, Fabienne Marcellin, Clemence Ramier, Camelia Protopopescu, Raphaelle Tardieu, Karine Ory, Dominique Salmon-Ceron, Patrizia Carrieri
JOURNAL OF VIRAL HEPATITIS
(2023)
Letter
Clinical Neurology
Tangui Barre, Damien Testa, Clemence Ramier, Melina Santos, Fabienne Marcellin, Perrine Roux, Patrizia Carrieri, Lise Radoszycki, Camelia Protopopescu
MULTIPLE SCLEROSIS AND RELATED DISORDERS
(2023)
Article
Gastroenterology & Hepatology
Fabienne Marcellin, Sylvie Bregigeon-Ronot, Clemence Ramier, Camelia Protopopescu, Camille Gilbert, Vincent Di Beo, Claudine Duvivier, Morgane Bureau-Stoltmann, Eric Rosenthal, Linda Wittkop, Dominique Salmon-Ceron, Patrizia Carrieri, Philippe Sogni, Tangui Barre
Summary: This study aimed to investigate the prevalence of moderate-to-severe depression in people living with HIV and HCV after successful HCV treatment, and identify associated socio-behavioral factors. Through descriptive and logistic regression analyses of data from 398 participants, it was found that 23.9% of HCV-cured individuals had moderate-to-severe depression. Female sex, unhealthy alcohol use, sedentary lifestyle, and unhealthy eating behaviors were associated with increased odds of moderate-to-severe depression.
Article
Psychology, Clinical
Davide Fortin, Vincent Di Beo, Patrizia Carrieri, Tangui Barre
Summary: This study explored the characteristics of French and Italian CBD users based on their primary reason for use. Users' characteristics differed based on their reasons to use CBD, but there were similarities between both countries for users who used CBD for therapeutic purposes. Analyzing the reasons for CBD use can help to characterize users and identify their unmet needs in terms of self-care.
JOURNAL OF PSYCHOACTIVE DRUGS
(2023)
Article
Pharmacology & Pharmacy
Tangui Barre, Vincenzo Di Marzo, Fabienne Marcellin, Patrizia Burra, Patrizia Carrieri
Summary: Obesity and nonalcoholic fatty liver disease (NAFLD) are global epidemics. The endocannabinoid system is seen as a key target for treating NAFLD. Although clinical trials are lacking, research suggests potential benefits of phytocannabinoids in treating liver steatosis. Cannabis plant should be considered a major prospect for NAFLD treatment. Overcoming scientific and non-scientific barriers is necessary for further research.
CANNABIS AND CANNABINOID RESEARCH
(2023)
Article
Pharmacology & Pharmacy
Tangui Barre, Fabrice Carrat, Clemence Ramier, Helene Fontaine, Vincent Di Beo, Morgane Bureau, Celine Dorival, Dominique Larrey, Elisabeth Delarocque-Astagneau, Philippe Mathurin, Fabienne Marcellin, Ventzislava Petrov-Sanchez, Carole Cagnot, Patrizia Carrieri, Stanislas Pol, Camelia Protopopescu
Summary: The use of cannabis is associated with smaller waist circumference, lower BMI, and lower risks of overweight, obesity, and central obesity in patients with chronic HCV infection. Longitudinal studies are needed to confirm these relationships and assess the impact of cannabis use on body weight and liver outcomes after HCV cure.
JOURNAL OF CANNABIS RESEARCH
(2022)