Article
Computer Science, Information Systems
Brahim Ait Skourt, Abdelhamid El Hassani, Aicha Majda
Summary: Deep neural networks are widely used in machine learning and regularization techniques, such as dropout, can effectively address the issue of overfitting. This study proposes a novel method called Mixed-Pooling-Dropout that combines dropout function with a mixed-pooling strategy, achieving better results than conventional pooling methods and max-pooling-dropout method.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Psychology, Developmental
Siman Liu, Yin-Shan Wang, Qing Zhang, Quan Zhou, Li-Zhi Cao, Chao Jiang, Zhe Zhang, Ning Yang, Qi Dong, Xi-Nian Zuo
Summary: The Chinese Color Nest Project (CCNP) aims to create normative charts for brain structure and function across the human lifespan and link age-related changes in brain imaging measures to psychological assessments. The project has recruited 1520 healthy individuals through three phases and collected data using an accelerated longitudinal design. The initial stage of data collection has been completed, with an emphasis on the developmental phase.
DEVELOPMENTAL COGNITIVE NEUROSCIENCE
(2021)
Article
Engineering, Marine
P. Romero-Tello, J. E. Gutierrez-Romero, B. Servan-Camas
Summary: The importance of predicting seakeeping in the early stages of ship design has led to the development of an artificial intelligence algorithm based on artificial neural networks. This AI tool aims to provide a fast and accurate prediction of seakeeping performance without requiring the exact hull shape.
JOURNAL OF OCEAN ENGINEERING AND SCIENCE
(2023)
Article
Gastroenterology & Hepatology
Dong Hyun Sinn, Danbee Kang, Mira Kang, Eliseo Guallar, Yun Soo Hong, Kyung Hyun Lee, Jiyeon Park, Juhee Cho, Geum-Youn Gwak
Summary: This cohort study found that patients with NAFLD are at increased risk of sarcopenia, as indicated by a faster loss of skeletal muscle mass. Patients with NAFLD, especially those with intermediate to high NFS, may require screening and early intervention to mitigate the loss of skeletal muscle mass.
Article
Biochemical Research Methods
Ying Zhang, Yuxin Song, Jin Gao, Hengyu Zhang, Ning Yang, Runqing Yang
Summary: The study extended the Hi-RRM model into a genome-wide association analysis for longitudinal data, reducing the dimensionality of repeated measurements significantly. The method improved computing efficiency through model transformation and demonstrated its statistical utility through simulation experiments.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Materials Science, Multidisciplinary
Menghuan Wang, Sheng Sun, Tong-Yi Zhang
Summary: This study utilizes machine learning methods to accelerate the design and optimization process of auxetic cellular structures. The generated initial data from finite element calculations are used to train an artificial neural network model, which demonstrates a search speed at least five orders of magnitude faster than the calculations themselves when exploring the structure space.
MATERIALS & DESIGN
(2023)
Article
Engineering, Mechanical
Hongju Li, Ying Liu, Haoran Liao, Zhurong Liang
Summary: This paper proposes an accelerated wear test design method based on entropy analysis of the dissipation wear model under mixed lubrication to achieve long-life predictions of tribo-pairs. The verification test shows that 20 times acceleration can be achieved by adjusting the entropy increase rate. The effect of the entropy increase rate gradient of duty parameters is also discussed.
Article
Education & Educational Research
Erin Murphy-Graham, Diana Pacheco Montoya, Alison K. Cohen, Enrique Valencia Lopez
Summary: The study found that dropout is a major problem among rural youth in Honduras, especially during transition years and the first year of upper secondary education. Dropout is the result of accumulation of push and pull factors, leading students to question the value of staying enrolled without opportunities to convert schooling into valued outcomes.
INTERNATIONAL JOURNAL OF EDUCATIONAL DEVELOPMENT
(2021)
Article
Mathematical & Computational Biology
Hongqian Wu, Michael P. Jones
Summary: The study proposes an extension of the generalized linear model for analyzing longitudinal data by incorporating random effects and using maximum likelihood for estimation and inference. Simulation studies are conducted to assess its properties and compare it to the generalized linear mixed model, with an illustration using data from a longitudinal study of Huntington's disease.
STATISTICS IN MEDICINE
(2021)
Article
Statistics & Probability
M. Taavoni, M. Arashi
Summary: This study proposes a semiparametric mixed effects model for analyzing CD4 longitudinal data, which uses a nonparametric function and random effects to handle time effects and within-subject correlations, and a parametric linear model for covariate effects. The research shows that, asymptotically, estimating regression coefficients using the profile kernel method is more accurate than the backfitting method, and thus, the use of the backfitting method is no longer recommended in semiparametric mixed effect longitudinal models.
STATISTICAL PAPERS
(2021)
Article
Engineering, Biomedical
Friederike Kaiser, Lena Schroeter, Svenja Stein, Benjamin Krueger, Jan Weichhold, Philipp Stahlhut, Anita Ignatius, Uwe Gbureck
Summary: Results of several studies have shown that magnesium phosphate cements (MPCs) have excellent biocompatibility, osteoconductivity, and higher solubility compared to calcium phosphates, making them promising alternatives to slowly degrading calcium phosphate cements. Decreasing the powder-liquid ratio (PLR) of the cement can further improve the degradation speed, with the potential for fast bone ingrowth within a few months. However, further modifications may be needed, especially for the K-struvite cement, to achieve a slower resorption and reduced inflammatory response in vivo.
ACTA BIOMATERIALIA
(2022)
Article
Education & Educational Research
Michelle Kaffenberger, Danielle Sobol, Debi Spindelman
Summary: Using longitudinal data, this study examines the role of learning in school persistence and dropout in Ethiopia, India, Peru, and Vietnam. The findings show a strong negative association between test scores and dropout rates, indicating that low learning plays a significant role in dropout decisions. Qualitative analysis further reveals that low learning exacerbates other causes of dropout, such as early marriage and leaving school to work.
INTERNATIONAL JOURNAL OF EDUCATIONAL RESEARCH
(2023)
Article
Medicine, General & Internal
Mohammad Inamul Haq, Thomas M. Drake, Tee Lin Goh, Asma Ahmed, Ewan Forrest, Stephen Barclay, Ruth Gillespie, Mathew Priest, Jeff Evans, Janet Graham, Stuart Ballantyne, Donald C. McMillan, Peter C. Hayes, Thomas G. Bird, Adrian J. Stanley
Summary: This study prospectively collected data on HCC patients in two large UK centres between January 2009 and December 2015, showing that adherence to surveillance was significantly associated with improved overall survival, even after adjusting for lead-time bias.
JOURNAL OF CLINICAL MEDICINE
(2021)
Article
Biochemistry & Molecular Biology
Maxwell L. Elliott, Daniel W. Belsky, Annchen R. Knodt, David Ireland, Tracy R. Melzer, Richie Poulton, Sandhya Ramrakha, Avshalom Caspi, Terrie E. Moffitt, Ahmad R. Hariri
Summary: BrainAGE, a measure of the difference between predicted age from brain-imaging data and chronological age, has been proposed as a biomarker for age-related brain deterioration. Studies suggest that individuals with older brainAGEs may be more susceptible to Alzheimer's, dementia, and mortality, but further research is needed to validate its effectiveness.
MOLECULAR PSYCHIATRY
(2021)
Article
Cardiac & Cardiovascular Systems
Charlotte Greve Sommer, Lars Bo Jorgensen, Birgitte Blume, Tom Moller, Soren Thorgaard Skou, Alexander Harrison, Lars Hermann Tang
Summary: This study aimed to investigate the dropout rate during a transitional exercise-based cardiac rehabilitation program and identify patient characteristics associated with dropout at the transition. The findings showed that the transition phase is critical in dropout prevention, emphasizing the importance of patient adherence and strategies for patient retention across different settings.
EUROPEAN JOURNAL OF CARDIOVASCULAR NURSING
(2022)