Article
Computer Science, Interdisciplinary Applications
Mingqiu Wang, Xiaoning Kang, Jiajuan Liang, Kun Wang, Yuanshan Wu
Summary: This paper proposes a method for simultaneous variable selection and heteroscedasticity identification for the linear location-scale model using a regularized multiple quantile regression approach. The method identifies heteroscedasticity, seeks common features of quantile coefficients, and eliminates irrelevant variables. Theoretical properties are established and simulation studies demonstrate the method's ability to identify covariates that affect the variability of the response variable. The method is also applied to analyse Wage data.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2023)
Article
Economics
Shuolin Shi, Ralf A. Wilke
Summary: This study examines the exit routes of older employees out of employment around retirement age, using administrative data and machine learning methods in a flexible regression model. The findings provide detailed insights into the impact of various administrative registers on early and late retirement transitions.
EMPIRICAL ECONOMICS
(2022)
Article
Economics
Zhanxiong Xu, Zhibiao Zhao
Summary: This study introduces an efficient estimator by constrainedly weighting information across quantiles, which can eliminate the effect of preliminary estimator and achieve good estimation efficiency simultaneously. Compared to the Cramer-Rao lower bound, the relative efficiency loss of the new estimator has a conservative upper bound close to zero in practical situations. Monte Carlo studies show that the proposed method has substantial efficiency gain and better prediction performance in empirical applications to GDP and inflation rate modeling.
JOURNAL OF BUSINESS & ECONOMIC STATISTICS
(2022)
Article
Economics
Jerry Hausman, Haoyang Liu, Ye Luo, Christopher Palmer
Summary: This study examines the impact of measurement error in random-coefficients models, specifically focusing on quantile regression. A sieve maximum likelihood approach is proposed to address the issue of left-hand-side measurement error, demonstrating consistency and asymptotic normality with appropriate conditions. Monte Carlo evidence indicates that this method outperforms traditional quantile regression in terms of bias and MSE estimation.
Article
Statistics & Probability
Thomas Nagler, Thibault Vatter
Summary: Copulas are commonly used to model complex dependence structures and can be applied to solve statistical learning problems. In this article, we generalize these approaches in a unified framework, allowing for simultaneous inferences across multiple regression-like problems. We also derive consistency, asymptotic normality, and validity of the bootstrap for the corresponding estimators. This methodology is versatile and can be applied to various types of data and estimators for copula and marginal distributions.
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
(2023)
Article
Mathematics, Applied
Yinfeng Wang, Xinsheng Zhang
Summary: This study examines quantile regression in linear models using the data-cutoff method, proposing consistent and asymptotically normal quantile estimators and applying adaptive LASSO for variable selection. The performance of the methods is evaluated through extensive numerical simulations, with foreign exchange rate data analyzed for illustration purposes.
SCIENCE CHINA-MATHEMATICS
(2022)
Article
Computer Science, Theory & Methods
Meadhbh O'Neill, Kevin Burke
Summary: Modern variable selection procedures utilize penalization methods to simultaneously perform model selection and estimation. The least absolute shrinkage and selection operator is a popular method that requires selecting a tuning parameter value. In contrast, our approach based on the smooth IC automatically selects the tuning parameter in one step. We also extend this method to the distributional regression framework, which offers more flexibility than classical regression modeling.
STATISTICS AND COMPUTING
(2023)
Article
Mathematics
Yongxia Zhang, Qi Wang, Maozai Tian
Summary: This paper investigates variable selection for a dataset with heavy-tailed distribution and high correlations within blocks of covariates. By introducing a latent factor model and a consistency strategy named Farvsqr, the study successfully addresses the challenges of high-dimensional data and highly correlated covariates.
Article
Statistics & Probability
Ye Fan, Nan Lin, Xianjun Yin
Summary: The paper introduces the QPADM-slack method, a parallel algorithm formulated via the ADMM that supports penalized quantile regression in big data, which is significantly faster and has favorable estimation accuracy compared to QPADM in large datasets.
JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS
(2021)
Article
Economics
Selahattin Murat Sirin, Berna N. Yilmaz
Summary: Global investment in renewable energy technologies is increasing, posing challenges to power markets. Studies show that as renewable energy generation increases, system prices decrease, but positive imbalances rise.
Article
Business, Finance
Maria Giuseppina Bruna, Rey Dang, Aymen Ammari, L'Hocine Houanti
Summary: This paper investigates the relationship between Women on Corporate Boards (WOCB) and Corporate Social Performance (CSP) using a sample of firms from the S&P 500 Index between 2004 and 2015. The study reveals a nonlinear threshold effect of board feminization on CSP, with changes along the quantiles of performance distribution. It contributes to corporate governance literature by highlighting the contextual and multilevel phenomenon of board dynamics and inclusiveness.
FINANCE RESEARCH LETTERS
(2021)
Article
Engineering, Civil
Qianlinglin Qiu, Zhongyao Liang, Yaoyang Xu, Shin-ichiro S. Matsuzaki, Kazuhiro Komatsu, Tyler Wagner
Summary: This study focuses on the temporal variation of CNRs and finds large interannual differences, with accumulative data being reliable for informing eutrophication management decisions in lakes. The novel statistical framework proposed serves as an important tool for estimating reliable CNRs and guiding lake-specific eutrophication control processes.
JOURNAL OF HYDROLOGY
(2021)
Article
Management
Sheng Dai
Summary: This paper proposes a new L-0-norm regularization approach to address the curse of dimensionality and validates its effectiveness through practical application and Monte Carlo study.
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH
(2023)
Article
Mathematics
Dengluan Dai, Anmin Tang, Jinli Ye
Summary: The high-dimensional quantile regression model improves computational efficiency and measures randomness through variational Bayesian methods and sparse penalty terms. Research shows that this method outperforms other regression methods in variable relationship studies.
Article
Mathematics
Yuanying Zhao, Dengke Xu
Summary: In this paper, a Bayesian variable selection method based on spike and slab prior for regression parameters is proposed for spatial autoregressive (SAR) quantile models. The proposed approach combines robust parametric estimation and variable selection in the context of SAR quantile models. Bayesian statistical inferences are implemented using a detailed Markov chain Monte Carlo (MCMC) procedure that combines Gibbs samplers with a probability integral transformation (PIT) algorithm. Empirical numerical examples, including simulation studies and a Boston housing price data analysis, are provided to demonstrate the newly developed methodologies.
Article
Behavioral Sciences
James A. Klarevas-Irby, Damien R. Farine
Summary: Little is known about how animals overcome temporal constraints on movement during dispersal. This study used GPS tracking of vulturine guineafowl and found that dispersers showed the greatest increase in movement at the same times of day when they moved the most prior to dispersal. These findings suggest that individuals face the same ecological constraints during dispersal as they do in daily life and achieve large displacements by maximizing movement when conditions are most favorable.
Article
Behavioral Sciences
Simone Ciaralli, Martina Esposito, Stefano Francesconi, Daniela Muzzicato, Marco Gamba, Matteo Dal Zotto, Daniela Campobello
Summary: Male cuckoos may transfer nest location information to females as a nonmaterial nuptial gift through specific postures and behaviors, potentially influencing mating choices of female cuckoos.
Article
Behavioral Sciences
Anne E. Aulsebrook, Rowan Jacques-Hamilton, Bart Kempenaers
Summary: Accelerometry and machine learning have been used to quantify mating behaviors of captive male ruffs. Different machine learning methods were compared and evaluated for their classification performance. The study highlights the challenges and potential pitfalls in classifying mating behaviors using accelerometry and provides recommendations and considerations for future research.
Article
Behavioral Sciences
Maria G. Smith, Joshua B. LaPergola, Christina Riehl
Summary: This study analyzed individual contributions to parental care in the greater ani bird and found that workload inequality varied between groups of two and three pairs. However, there was no clear evidence of division of labour within the groups, suggesting individual differences in overall work performed.
Article
Behavioral Sciences
Noah M. T. Smith, Reuven Dukas
Summary: Winner and loser effects are observed in many animals, and recent experiments suggest that they may also occur in humans. In two experiments involving video games and reading comprehension, participants who won in the first phase performed significantly better in the second phase compared to those who lost. The effect size was larger in the video game experiment, and men and women showed similar magnitudes of winner and loser effects.
Article
Behavioral Sciences
Bianca J. L. Marcellino, Peri Yee, Shannon J. Mccauley, Rosalind L. Murray
Summary: This study examines the trade-off between mating effort and thermoregulatory behavior in dragonflies in response to temperature changes, and investigates the effect of wing melanin on these behaviors. The results indicate that as temperature increases, dragonflies reduce their mating effort and increase their thermoregulatory behavior.
Article
Behavioral Sciences
Rafael Rios Moura, Paulo Inacio Prado, Joao Vasconcellos-Neto
Summary: This study examined the escape behavior and decision-making of Aglaoctenus castaneus spiders on different substrates. It was found that spiders inhabiting injurious substrates displayed shorter flight initiation distances and lower sensitivity to predators.
Article
Behavioral Sciences
Luigi Baciadonna, Cwyn Solvi, Francesca Terranova, Camilla Godi, Cristina Pilenga, Livio Favaro
Summary: In this study, it was found that African penguins could use ventral dot patterns to recognize their lifelong partner and nonpartner colonymates. This challenges the previous assumption of limited visual involvement in penguin communication, highlighting the complex and flexible recognition process in birds.
Article
Behavioral Sciences
Nick A. R. Jones, Jade Newton-Youens, Joachim G. Frommen
Summary: Environmental conditions, particularly temperature, have a significant impact on animal behavior. This study focused on aggression in Neolamprologus pulcher fish and found that aggression rates increased with temperature at lower levels, but decreased after reaching a peak. Additionally, the influence of high temperatures on aggression changed over time during the trials. These findings provide a more comprehensive understanding of the short-term effects of temperature on aggression and highlight the importance of considering non-linear changes in thermal performance.
Article
Behavioral Sciences
Bruno Herlander Martins, Andrea Soriano-Redondo, Aldina M. A. Franco, Ines Carry
Summary: Human activities have affected the availability of resources for wildlife, particularly through the provision of anthropogenic food subsidies at landfill sites. This study explores the influence of age on landfill attendance and foraging behavior in white storks. Adult storks visit landfills more frequently and show dominance over juveniles in food acquisition. Juveniles have limited access to landfill resources and are forced to use lower quality areas.