Article
Health Care Sciences & Services
Glen P. Martin, Richard D. Riley, Gary S. Collins, Matthew Sperrin
Summary: Recent research suggests that penalisation methods should be used in the development of clinical prediction models to further reduce overfitting, although this may lead to increased variability in predictive performance in external data. It is recommended to always evaluate the variability in predictive performance and tuning parameters to ensure the reliability of the developed model.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2021)
Article
Health Care Sciences & Services
Richard D. Riley, Kym I. E. Snell, Glen P. Martin, Rebecca Whittle, Lucinda Archer, Matthew Sperrin, Gary S. Collins
Summary: When developing a clinical prediction model, penalization techniques may be unreliable due to large uncertainty in estimated tuning parameters, especially with small effective sample sizes and low model overfitting levels. This uncertainty can result in considerable miscalibration of model predictions for new individuals.
JOURNAL OF CLINICAL EPIDEMIOLOGY
(2021)
Article
Public, Environmental & Occupational Health
Pragyan Monalisa Sahoo, Himanshu Sekhar Rout, Mihajlo Jakovljevic
Summary: This study examined the health expenditure trend among the BRICS countries from 2000 to 2019 and made predictions for 2035 with a focus on public, pre-paid, and out-of-pocket expenditures. Except for India and Brazil, all BRICS countries show a long-term increase in per capita PPP health expenditure. India's health expenditure is expected to decrease as a share of GDP after the completion of the SDG years. China has the highest per capita expenditure growth rate, while Russia is expected to achieve the highest absolute values. The BRICS countries have the potential to be important leaders in various social policies such as health, and their estimations of future health expenditures can help policymakers allocate resources to achieve universal health coverage.
GLOBALIZATION AND HEALTH
(2023)
Article
Computer Science, Information Systems
Ioannis E. Livieris, Niki Kiriakidou, Stavros Stavroyiannis, Panagiotis Pintelas
Summary: Cryptocurrencies are widely recognized as an alternative exchange currency method due to their significant volatility and price fluctuations. The development of an accurate and reliable forecasting model is essential for cryptocurrency portfolio management and optimization. A multiple-input deep neural network model has been proposed for predicting cryptocurrency price and movement, showing the ability to efficiently utilize mixed cryptocurrency data.
Article
Economics
Qin Luo, Feng Ma, Jiqian Wang, You Wu
Summary: Academic research uses exogenous drivers to improve the accuracy of oil volatility forecasting. Dimension reduction regressions, particularly principal component analysis regression, successfully predict oil volatility at the one-month horizon. Shrinkage methods outperform other methods for medium and long-term forecast horizons. Unsupervised learning (PCA) performs better during periods of oil price decrease, while supervised learning methods (shrinkage methods) significantly improve volatility accuracy. The empirical results also identify several economic indicators that have a significant impact on oil volatility.
Article
Green & Sustainable Science & Technology
Tetiana Zatonatska, Olena Liashenko, Yana Fareniuk, Oleksandr Dluhopolskyi, Artur Dmowski, Marzena Cichorzewska
Summary: This paper forecasts the level of health care expenditure in Ukraine for 2023-2024 under different scenarios, considering the impact of migration and GDP decline. It highlights the need for prompt monitoring and adjustments to health care costs due to population aging and significant migration caused by war.
Article
Hospitality, Leisure, Sport & Tourism
Yishuo Zhang, Gang Li, Birgit Muskat, Rob Law
Summary: The study improves the accuracy of AI-based tourism demand forecasting models using a decomposition method, addressing overfitting issues.
JOURNAL OF TRAVEL RESEARCH
(2021)
Article
Engineering, Multidisciplinary
Mauricio Pereira, Branko Glisic
Summary: This article investigates the calibration strategies of creep and shrinkage models using SHM data for data-driven forecasting of long-term time-dependent behavior of high-rise buildings. A calibration strategy is identified that significantly improves forecasting and provides accurate predictions at least 30 days ahead.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2023)
Article
Engineering, Civil
S. D. D. A. Gedara, P. L. P. Wasantha, B. Teodosio, J. Li
Summary: This study investigates the influence of specimen size on axial and radial strains in core shrinkage tests. It found that larger specimens showed lower strains, attributing this behavior to the presence of more shrinkage cracks. Additionally, the axial and volumetric strains did not show significant variations with different length-to-diameter ratios.
TRANSPORTATION GEOTECHNICS
(2022)
Article
Public, Environmental & Occupational Health
Guvenc Kockaya, Gulpembe Oguzhan, Zafer Calskan
Summary: Since the implementation of the Health Transformation Program in Turkey in 2003, significant changes have been seen in the healthcare system. However, the percentage of households with catastrophic health expenditures has started to rise since 2012. The study evaluates the expenditure items that may have caused this increase, finding that the decrease in medicine expenditures has had a positive impact on household health spending.
FRONTIERS IN PUBLIC HEALTH
(2021)
Article
Public, Environmental & Occupational Health
Wangzi Xu, Jia Lin
Summary: Fiscal decentralization has a positive effect on local public health expenditure and improves regional public health through increased spending and improved availability of medical services.
FRONTIERS IN PUBLIC HEALTH
(2022)
Article
Endocrinology & Metabolism
Andrew W. McHill, Lindsey S. Brown, Andrew J. K. Phillips, Laura K. Barger, Marta Garaulet, Frank A. J. L. Scheer, Elizabeth B. Klerman
Summary: This study examined the relationship between the timing of energy intake and body composition based on mathematically modeled circadian timing and in-laboratory collected metrics. The findings suggest that the use of mathematically modeled circadian timing can provide similar results to in-laboratory measurements, which may be beneficial in time-based interventions.
Article
Computer Science, Artificial Intelligence
Hyun Jun Park, Youngjun Kim, Ha Young Kim
Summary: This study proposes a novel stock market prediction framework, LSTM-Forest, which integrates long short-term memory and random forest to address the issue of overfitting due to an increase in input variables. A multi-task model is developed to predict stock market returns and classify return directions, with interpretability achieved through variable importance analysis from random forest. Experimental results show that LFM outperforms baseline models and single-task models in predicting returns and return directions.
APPLIED SOFT COMPUTING
(2022)
Article
Economics
Peter Nystrup, Erik Lindstrom, Jan K. Moller, Henrik Madsen
Summary: A method utilizing eigendecomposition for dimensionality reduction in reconciling forecasts is proposed to improve accuracy and applicability, being applicable to hierarchies of all sizes.
INTERNATIONAL JOURNAL OF FORECASTING
(2021)
Article
Energy & Fuels
Hjorleifur G. Bergsteinsson, Jan Kloppenborg Moller, Peter Nystrup, Olafur Petur Palsson, Daniela Guericke, Henrik Madsen
Summary: This paper proposes a method to enhance the accuracy of heat load forecasts in district heating by using temporal hierarchies to coordinate forecasts from multiple aggregation levels, ultimately achieving more accurate predictions.
Article
Economics
Terry N. Flynn, Marcel Bilger, Chetna Malhotra, Eric A. Finkelstein
Article
Economics
Marcel Bilger, Eliza J. Kruger, Eric A. Finkelstein
Article
Medicine, Research & Experimental
Marcel Bilger, Tina T. Wong, Kaye L. Howard, Jia Yi Lee, Ai Nee Toh, Geraldine John, Ecosse L. Lamoureux, Eric A. Finkelstein
Article
Economics
Di Dong, Semra Ozdemir, Yong Mong Bee, Sue-Anne Toh, Marcel Bilger, Eric Finkelstein
Article
Endocrinology & Metabolism
Eric A. Finkelstein, Benjamin A. Haaland, Marcel Bilger, Aarti Sahasranaman, Robert A. Sloan, Ei Ei Khaing Nang, Kelly R. Evenson
LANCET DIABETES & ENDOCRINOLOGY
(2016)
Article
Health Care Sciences & Services
Wan Chen Kang Graham, Marcel Bilger
Article
Medicine, Research & Experimental
Marcel Bilger, Mitesh Shah, Ngiap Chuan Tan, Kaye Louise Howard, Hui Yan Xu, Ecosse Luc Lamoureux, Eric Andrew Finkelstein
Article
Oncology
Pui San Tan, Marcel Bilger, Gilberto de Lima Lopes, Sanchalika Acharyya, Benjamin Haaland
Article
Economics
Semra Ozdemir, Marcel Bilger, Eric A. Finkelstein
APPLIED HEALTH ECONOMICS AND HEALTH POLICY
(2017)
Article
Medicine, Research & Experimental
Yu Heng Kwan, Warren Fong, Xiang Ling Ang, Chuen Seng Tan, Bee Choo Tai, Youyi Huang, Marcel Bilger, Jie Kie Phang, Hui Chin Tan, Jia Ven Lee, Limin Sun, Choy Tip Tan, Bao Qiang Dong, Hwee Ling Koh, Ying Ying Leung, Nai Lee Lui, Siaw Ing Yeo, Swee Cheng Ng, Kok Yong Fong, Julian Thumboo, Truls Ostbye
Review
Public, Environmental & Occupational Health
Eric A. Finkelstein, Marcel Bilger, Drishti Baid
SOCIAL SCIENCE & MEDICINE
(2019)
Article
Economics
Marcel Bilger, Tina T. Wong, Jia Yi Lee, Kaye L. Howard, Filipinas G. Bundoc, Ecosse L. Lamoureux, Eric A. Finkelstein
APPLIED HEALTH ECONOMICS AND HEALTH POLICY
(2019)
Article
Health Care Sciences & Services
Marcel Bilger, Agnes Ying Leng Koong, Ian Kwong Yun Phoon, Ngiap Chuan Tan, Juliana Bahadin, Joann Bairavi, Ada Portia M. Batcagan-Abueg, Eric A. Finkelstein
Summary: This study aimed to investigate whether wireless home blood pressure monitoring with or without financial incentives is more effective at reducing systolic blood pressure in patients with hypertension. The study was designed as a randomized controlled trial, but it was stopped prematurely due to unforeseen events, therefore no results are available at this time.
JMIR RESEARCH PROTOCOLS
(2021)
Article
Health Care Sciences & Services
Marcel Bilger, Mitesh Shah, Ngiap Chuan Tan, Cynthia Y. L. Tan, Filipinas G. Bundoc, Joann Bairavi, Eric A. Finkelstein
Summary: The study found that incorporating financial incentives into diabetes self-management can help improve glycemic control among patients. Process-based incentives are more effective than outcome-based incentives in promoting medication adherence and physical activity.
PATIENT-PATIENT CENTERED OUTCOMES RESEARCH
(2021)