In clinical research,subgroup analysis can help identify patient groups that respond better or worse to specific treatments,improve therapeutic effect and safety,and is of great significance in precision medicine.This...In clinical research,subgroup analysis can help identify patient groups that respond better or worse to specific treatments,improve therapeutic effect and safety,and is of great significance in precision medicine.This article considers subgroup analysis methods for longitudinal data containing multiple covariates and biomarkers.We divide subgroups based on whether a linear combination of these biomarkers exceeds a predetermined threshold,and assess the heterogeneity of treatment effects across subgroups using the interaction between subgroups and exposure variables.Quantile regression is used to better characterize the global distribution of the response variable and sparsity penalties are imposed to achieve variable selection of covariates and biomarkers.The effectiveness of our proposed methodology for both variable selection and parameter estimation is verified through random simulations.Finally,we demonstrate the application of this method by analyzing data from the PA.3 trial,further illustrating the practicality of the method proposed in this paper.展开更多
Branch size is a crucial characteristic,closely linked to both tree growth and wood quality.A review of existing branch size models reveals various approaches,but the ability to estimate branch diameter and length wit...Branch size is a crucial characteristic,closely linked to both tree growth and wood quality.A review of existing branch size models reveals various approaches,but the ability to estimate branch diameter and length within the same whorl remains underexplored.In this study,a total of 77 trees were sampled from Northeast China to model the vertical distribution of branch diameter and length within each whorl along the crown.Several commonly used functions were taken as the alternative model forms,and the quantile regression method was employed and compared with the classical two-step modeling approach.The analysis incorporated stand,tree,and competition factors,with a particular focus on how these factors influence branches of varying sizes.The modified Weibull function was chosen as the optimal model,due to its excellent performance across all quantiles.Eight quantile regression curves(ranging from 0.20 to 0.85)were combined to predict branch diameter,while seven curves(ranging from 0.20 to 0.80)were used for branch length.The results showed that the quantile regression method outperformed the classical approach at model fitting and validation,likely due to its ability to estimate different rates of change across the entire branch size distribution.Lager branches in each whorl were more sensitive to changes in DBH,crown length(CL),crown ratio(CR)and dominant tree height(H_(dom)),while slenderness(HDR)more effectively influenced small and medium-sized branches.The effect of stand basal area(BAS)was relatively consistent across different branch sizes.The findings indicate that quantile regression is a good way not only a more accurate method for predicting branch size but also a valuable tool for understanding how branch growth responds to stand and tree factors.The models developed in this study are prepared to be further integrated into tree growth and yield simulation system,contributing to the assessment and promotion of wood quality.展开更多
Despite the maturity of ensemble numerical weather prediction(NWP),the resulting forecasts are still,more often than not,under-dispersed.As such,forecast calibration tools have become popular.Among those tools,quantil...Despite the maturity of ensemble numerical weather prediction(NWP),the resulting forecasts are still,more often than not,under-dispersed.As such,forecast calibration tools have become popular.Among those tools,quantile regression(QR)is highly competitive in terms of both flexibility and predictive performance.Nevertheless,a long-standing problem of QR is quantile crossing,which greatly limits the interpretability of QR-calibrated forecasts.On this point,this study proposes a non-crossing quantile regression neural network(NCQRNN),for calibrating ensemble NWP forecasts into a set of reliable quantile forecasts without crossing.The overarching design principle of NCQRNN is to add on top of the conventional QRNN structure another hidden layer,which imposes a non-decreasing mapping between the combined output from nodes of the last hidden layer to the nodes of the output layer,through a triangular weight matrix with positive entries.The empirical part of the work considers a solar irradiance case study,in which four years of ensemble irradiance forecasts at seven locations,issued by the European Centre for Medium-Range Weather Forecasts,are calibrated via NCQRNN,as well as via an eclectic mix of benchmarking models,ranging from the naïve climatology to the state-of-the-art deep-learning and other non-crossing models.Formal and stringent forecast verification suggests that the forecasts post-processed via NCQRNN attain the maximum sharpness subject to calibration,amongst all competitors.Furthermore,the proposed conception to resolve quantile crossing is remarkably simple yet general,and thus has broad applicability as it can be integrated with many shallow-and deep-learning-based neural networks.展开更多
This paper introduces a novel approach for parameter sensitivity evaluation and efficient slope reliability analysis based on quantile-based first-order second-moment method(QFOSM).The core principles of the QFOSM are...This paper introduces a novel approach for parameter sensitivity evaluation and efficient slope reliability analysis based on quantile-based first-order second-moment method(QFOSM).The core principles of the QFOSM are elucidated geometrically from the perspective of expanding ellipsoids.Based on this geometric interpretation,the QFOSM is further extended to estimate sensitivity indices and assess the significance of various uncertain parameters involved in the slope system.The proposed method has the advantage of computational simplicity,akin to the conventional first-order second-moment method(FOSM),while providing estimation accuracy close to that of the first-order reliability method(FORM).Its performance is demonstrated with a numerical example and three slope examples.The results show that the proposed method can efficiently estimate the slope reliability and simultaneously evaluate the sensitivity of the uncertain parameters.The proposed method does not involve complex optimization or iteration required by the FORM.It can provide a valuable complement to the existing approximate reliability analysis methods,offering rapid sensitivity evaluation and slope reliability analysis.展开更多
Copula functions have been widely used in stochastic simulation and prediction of streamflow.However,existing models are usually limited to single two-dimensional or three-dimensional copulas with the same bivariate b...Copula functions have been widely used in stochastic simulation and prediction of streamflow.However,existing models are usually limited to single two-dimensional or three-dimensional copulas with the same bivariate block for all months.To address this limitation,this study developed a mixed D-vine copula-based conditional quantile model that can capture temporal correlations.This model can generate streamflow by selecting different historical streamflow variables as the conditions for different months and by exploiting the conditional quantile functions of streamflows in different months with mixed D-vine copulas.The up-to-down sequential method,which couples the maximum weight approach with the Akaike information criteria and the maximum likelihood approach,was used to determine the structures of multivariate Dvine copulas.The developed model was used in a case study to synthesize the monthly streamflow at the Tangnaihai hydrological station,the inflow control station of the Longyangxia Reservoir in the Yellow River Basin.The results showed that the developed model outperformed the commonly used bivariate copula model in terms of the performance in simulating the seasonality and interannual variability of streamflow.This model provides useful information for water-related natural hazard risk assessment and integrated water resources management and utilization.展开更多
Purpose–This study aimed to facilitate a rapid evaluation of track service status and vehicle ride comfort based on car body acceleration.Consequently,a low-cost,data-driven approach was proposed for analyzing speed-...Purpose–This study aimed to facilitate a rapid evaluation of track service status and vehicle ride comfort based on car body acceleration.Consequently,a low-cost,data-driven approach was proposed for analyzing speed-related acceleration limits in metro systems.Design/methodology/approach–A portable sensing terminal was developed to realize easy and efficient detection of car body acceleration.Further,field measurements were performed on a 51.95-km metro line.Data from 272 metro sections were tested as a case study,and a quantile regression method was proposed to fit the control limits of the car body acceleration at different speeds using the measured data.Findings–First,the frequency statistics of the measured data in the speed-acceleration dimension indicated that the car body acceleration was primarily concentrated within the constant speed stage,particularly at speeds of 15.4,18.3,and 20.9 m/s.Second,resampling was performed according to the probability density distribution of car body acceleration for different speed domains to achieve data balance.Finally,combined with the traditional linear relationship between speed and acceleration,the statistical relationships between the speed and car body acceleration under different quantiles were determined.We concluded the lateral/vertical quantiles of 0.8989/0.9895,0.9942/0.997,and 0.9998/0.993 as being excellent,good,and qualified control limits,respectively,for the lateral and vertical acceleration of the car body.In addition,regression lines for the speedrelated acceleration limits at other quantiles(0.5,0.75,2s,and 3s)were obtained.Originality/value–The proposed method is expected to serve as a reference for further studies on speedrelated acceleration limits in rail transit systems.展开更多
In this paper,we introduce the censored composite conditional quantile coefficient(cC-CQC)to rank the relative importance of each predictor in high-dimensional censored regression.The cCCQC takes advantage of all usef...In this paper,we introduce the censored composite conditional quantile coefficient(cC-CQC)to rank the relative importance of each predictor in high-dimensional censored regression.The cCCQC takes advantage of all useful information across quantiles and can detect nonlinear effects including interactions and heterogeneity,effectively.Furthermore,the proposed screening method based on cCCQC is robust to the existence of outliers and enjoys the sure screening property.Simulation results demonstrate that the proposed method performs competitively on survival datasets of high-dimensional predictors,particularly when the variables are highly correlated.展开更多
This paper aims to explore the application of Extreme Value Theory (EVT) in estimating the conditional extreme quantile for time-to-event outcomes by examining the functional relationship between ambulatory blood pres...This paper aims to explore the application of Extreme Value Theory (EVT) in estimating the conditional extreme quantile for time-to-event outcomes by examining the functional relationship between ambulatory blood pressure trajectories and clinical outcomes in stroke patients. The study utilizes EVT to analyze the functional connection between ambulatory blood pressure trajectories and clinical outcomes in a sample of 297 stroke patients. The 24-hour ambulatory blood pressure measurement curves for every 15 minutes are considered, acknowledging a censored rate of 40%. The findings reveal that the sample mean excess function exhibits a positive gradient above a specific threshold, confirming the heavy-tailed distribution of data in stroke patients with a positive extreme value index. Consequently, the estimated conditional extreme quantile indicates that stroke patients with higher blood pressure measurements face an elevated risk of recurrent stroke occurrence at an early stage. This research contributes to the understanding of the relationship between ambulatory blood pressure and recurrent stroke, providing valuable insights for clinical considerations and potential interventions in stroke management.展开更多
As extremely important methods, Lp regression methods have attracted the attention of either theoretical or empirical researchers all over the world. As special cases of that, quantile and expectile regression (with p...As extremely important methods, Lp regression methods have attracted the attention of either theoretical or empirical researchers all over the world. As special cases of that, quantile and expectile regression (with p = 1 and p = 2, respectively) are well acquainted with. In contrast with them, general Lp regression (with p equals 1 and 2) has recently been found to have many unexpected properties by some studies, especially when 1 p Lp quantile regression under various p settings and shows some recent advances in Lp quantile regression, theoretically and empirically.展开更多
In the context of the rapidly growing demand for green investments and the need to combat climate change,this study contributes to the emerging literature on green investments by exploring the time-frequency connected...In the context of the rapidly growing demand for green investments and the need to combat climate change,this study contributes to the emerging literature on green investments by exploring the time-frequency connectedness between green bonds(GBs)and green equities.Specifically,we examine the degree of connection between GBs and green equities,the extent to which these markets influence each other,and which one is the primary net transmitter versus the net receiver of shocks under diverse market conditions.To accomplish these objectives,we use the wavelet-based Quantile-on-Quantile(QQ),dynamic conditional correlation(DCC),portfolio implications,and Quantile VAR approaches.The results show that GBs and green equities have a strong positive connection,depending on time and frequency domains.However,a negative association between GBs and green equities is observed during periods of crisis,highlighting GBs’ability to hedge green equity portfolios.The portfolio strategies demonstrate that investors require to invest in the Green Economy equity and S&P GB portfolio to reach the highest level of hedging effectiveness.The findings further imply that the Global Water Equity Index transmits the highest spillover to other green assets,while the Green Economy Equity Index receives the most spillover from other assets.The pairwise volatility connectivity reveals that most pairs have minimal quantile dependence,indicating the potential for diversification across the GB and green equity pairs.These findings have significant implications for investors and policymakers concerned with green investments and climate change mitigation.展开更多
Value at risk(VaR)and expected shortfall(ES)have emerged as standard measures for detecting the market risk of financial assets and play essential roles in investment decisions,external regulations,and risk capital al...Value at risk(VaR)and expected shortfall(ES)have emerged as standard measures for detecting the market risk of financial assets and play essential roles in investment decisions,external regulations,and risk capital allocation.However,existing VaR estimation approaches fail to accurately reflect downside risks,and the ES estimation technique is quite limited owing to its challenging implementation.This causes financial institutions to overestimate or underestimate investment risk and finally leads to the inefficient allocation of financial resources.The main purpose of this study is to use machine learning to improve the accuracy of VaR estimation and provide an effective tool for ES estimation.Specifically,this study proposes a VaR estimator by combining quantile regression with“Mogrifier”recurrent neural networks to capture the“long memory”and“clustering”properties of financial assets;while for estimating ES,this study directly models the quantile of assets and employs generative adversarial networks to generate future tail risk scenarios.In addition to the typical properties of financial assets,the model design is also consistent with heterogeneous market theory.An empirical application to four major global stock indices shows that our model is superior to other existing models.展开更多
There is a substantial body of empirical research that has found the fund return distributions to exhibit pronounced peakiness, heavy tails, and skewness, deviating from a normal distribution. Addressing the limitatio...There is a substantial body of empirical research that has found the fund return distributions to exhibit pronounced peakiness, heavy tails, and skewness, deviating from a normal distribution. Addressing the limitations of the traditional Sharpe ratio, which assumes a normal distribution of returns and uses standard deviation to measure investment risk, this paper primarily employs the Value at Risk (VaR) based on Lp quantile to adjust excess returns of funds. This method offers superior robustness, is capable of capturing asymmetry and heavy-tailed characteristics, and is more flexible, providing a better description of the tail risk in fund returns. Empirical studies have shown that using the Sharpe ratio corrected with the Lp quantile is feasible for evaluating and ranking the performance of open-end funds.展开更多
Analyzing the interdependencies among financial institutions is critical for designing systemic risk monitoring mechanisms;however,most existing research focuses on the first moment of the return distribution,which fa...Analyzing the interdependencies among financial institutions is critical for designing systemic risk monitoring mechanisms;however,most existing research focuses on the first moment of the return distribution,which falls into the conventional models of choice under risk.Previous literature has observed the scarcity of investors’attention and processing power,which makes the traditional theory of choice under risk more vulnerable and brings the salience theory that accommodates investors’cognitive limitations to our attention.Motivated by evidence of salience theory value(STV)containing unique information not captured by traditional higher-order moments,we employ a quantile connectedness approach to examine the STV interconnectedness of China’s systemically important banks(C-SIBs).The quantile approach allows us to uncover the dynamic STV interconnectedness of C-SIBs under normal,bearish,and bullish market conditions and is well-suited to extreme risk problems.Our results show that the C-SIBs system is asymmetrically interconnected across quantiles and at higher levels under bullish than bearish market conditions.Principally,a bank’s performance in the C-SIBs system depends on its systemic importance and market conditions.Furthermore,the comparative analysis indicates that STV could provide more information than higher-order moments in capturing the dynamic change in the C-SIBs system and detecting some market events more precisely.These results have important implications for policymakers and market participants to formulate regulatory policy and design risk management strategies.展开更多
This article introduces a three-parameter Lehman-type t distribution having 2 degrees of freedom,that is capable of fitting positive and negative skewed data sets.It is shown that the density and hazard functions of t...This article introduces a three-parameter Lehman-type t distribution having 2 degrees of freedom,that is capable of fitting positive and negative skewed data sets.It is shown that the density and hazard functions of the proposed distribution are uni-model.Ordinary moments,entropy measure,ordering,identifiability and order statistics are investigated.Since the quantile function is explicitly defined,quantile-based statistics are also discussed for the proposed distribution.These properties include measures of skewness and kurtosis,L-moments,quantile density and hazard functions,mean residual life function and Parzen's score function.Mechanisms of maximum likelihood,bias correction and matching of percentiles are employed for estimating the unknown parameters of the distribution.Simulation experiments are conducted to compare the performance of these three estimation methods.A real-life data set consisting of strength of glass fibres is fitted to show the adequacy of the proposed distribution over some extensions of the normal and t distributions.Parametric regression model is developed along with its parameter estimation using the maximum likelihood approach.Simulation study for the regression model is also presented that endorsed the asymptotic properties of the estimators.展开更多
Since its inception,the epsilon distribution has piqued the interest of statisticians.It has been successfully used to solve a variety of statistical problems.In this article,we propose to use the quadratic rank trans...Since its inception,the epsilon distribution has piqued the interest of statisticians.It has been successfully used to solve a variety of statistical problems.In this article,we propose to use the quadratic rank transmutation map mechanism to extend this distribution.This mechanism is not new;it was already used to improve the modeling capabilities of a number of existing distributions.For the original epsilon distribution,we expect the same benefits.As a result,we implement the transmuted epsilon distribution as a flexible three-parameter distribution with a bounded domain.We demonstrate its key features,focusing on the properties of its distributional mechanism and conducting quantile and moment analyses.Applications of the model are presented using two data sets.We also perform a regression analysis based on this distribution.展开更多
Heatwaves are becoming increasingly frequent and severe,posing escalating risks to ecosystems and human well-being.While soil moisture(SM)deficits are recognized as important contributors to heatwave amplification,the...Heatwaves are becoming increasingly frequent and severe,posing escalating risks to ecosystems and human well-being.While soil moisture(SM)deficits are recognized as important contributors to heatwave amplification,their spatially heterogeneous impacts across the Northern Hemisphere remain insufficiently understood.In this study,we analyze ERA5 reanalysis data(1980-2022)to investigate trends in heatwave frequency,intensity,and duration,as well as their sensitivity to SM variability.Our results show robust increases in heatwave occurrence(0.76 events per decade),intensity(0.81℃per decade),and average duration(0.40 days per decade),with extreme events,as represented by maximum intensity and duration,rising at even faster rates(2.18℃per decade and 0.83 days per decade,respectively).Strong negative correlations are observed between SM deficits and heatwave metrics,with the magnitude of this relationship varying across land cover types and heatwave severity levels.Quantile regression reveals that SM reductions have a greater impact at higher quantiles for most indicators.Cropland exhibits the highest sensitivity to SM anomalies,whereas forests show more resilience due to their superior water retention capacities.These findings underscore the crucial role of land-atmosphere interactions in shaping heatwave extremes,providing a scientific basis for enhancing early warning and adaptation strategies in the context of ongoing climate change.展开更多
As the world’s fourth most populous country,Indonesia presents challenges and opportunities for sustainable energy progress,offering a critical context to investigate green human development(GHD).This study uniquely ...As the world’s fourth most populous country,Indonesia presents challenges and opportunities for sustainable energy progress,offering a critical context to investigate green human development(GHD).This study uniquely contributes to the literature by employing the planetary pressures-adjusted human development index(PHDI)as an indicator of GHD,which integrates environmental impacts into human development.Using static and dynamic econometric methods,including the quantile regression and autoregressive distributed lag model,it explores the impacts of renewable and nonrenewable energy consumption on GHD.The findings demonstrate that renewable energy currently has a detrimental impact on GHD due to its limited adoption and high costs.Conversely,nonrenewable energy positively influences GHD,as it is the primary energy source in the country and is becoming more efficient at reducing emissions.However,the study finds that greater use of renewable energy reduces its adverse effects,suggesting that as renewable energy technologies become more cost-effective and widely implemented,their initial adverse effects could be mitigated,leading to improved long-term GHD outcomes.These findings carry important implications for Indonesia,where the govern‐ment is striving to expand renewable energy capacity while promoting equitable development across its archi‐pelagic regions.They underscore the critical role of energy policy in balancing economic,social,and environmental goals,contributing meaningfully to the country’s sustainable development agenda.展开更多
Accurate forecasting of oil production is essential for optimizing resource management and minimizing operational risks in the energy sector. Traditional time-series forecasting techniques, despite their widespread ap...Accurate forecasting of oil production is essential for optimizing resource management and minimizing operational risks in the energy sector. Traditional time-series forecasting techniques, despite their widespread application, often encounter difficulties in handling the complexities of oil production data, which is characterized by non-linear patterns, skewed distributions, and the presence of outliers. To overcome these limitations, deep learning methods have emerged as more robust alternatives. However, while deep neural networks offer improved accuracy, they demand substantial amounts of data for effective training. Conversely, shallow networks with fewer layers lack the capacity to model complex data distributions adequately. To address these challenges, this study introduces a novel hybrid model called Transfer LSTM to GRU (TLTG), which combines the strengths of deep and shallow networks using transfer learning. The TLTG model integrates Long Short-Term Memory (LSTM) networks and Gated Recurrent Units (GRU) to enhance predictive accuracy while maintaining computational efficiency. Gaussian transformation is applied to the input data to reduce outliers and skewness, creating a more normal-like distribution. The proposed approach is validated on datasets from various wells in the Tahe oil field, China. Experimental results highlight the superior performance of the TLTG model, achieving 100% accuracy and faster prediction times (200 s) compared to eight other approaches, demonstrating its effectiveness and efficiency.展开更多
基金Supported by the Natural Science Foundation of Fujian Province(2022J011177,2024J01903)the Key Project of Fujian Provincial Education Department(JZ230054)。
文摘In clinical research,subgroup analysis can help identify patient groups that respond better or worse to specific treatments,improve therapeutic effect and safety,and is of great significance in precision medicine.This article considers subgroup analysis methods for longitudinal data containing multiple covariates and biomarkers.We divide subgroups based on whether a linear combination of these biomarkers exceeds a predetermined threshold,and assess the heterogeneity of treatment effects across subgroups using the interaction between subgroups and exposure variables.Quantile regression is used to better characterize the global distribution of the response variable and sparsity penalties are imposed to achieve variable selection of covariates and biomarkers.The effectiveness of our proposed methodology for both variable selection and parameter estimation is verified through random simulations.Finally,we demonstrate the application of this method by analyzing data from the PA.3 trial,further illustrating the practicality of the method proposed in this paper.
基金supported by the Young Scientists Fund of the National Key R&D Program of China(No.2022YFD2201800)the Youth Science Fund Program of National Natural Science Foundation of China(No.32301581)+2 种基金the Joint Funds for Regional Innovation and Development of the National Natural Science Foundation of China(No.U21A20244)the China Postdoctoral Science Foundation(No.2024M750383)the Heilongjiang Touyan Innovation Team Program(Technology Development Team for High-Efficiency Silviculture of Forest Resources).
文摘Branch size is a crucial characteristic,closely linked to both tree growth and wood quality.A review of existing branch size models reveals various approaches,but the ability to estimate branch diameter and length within the same whorl remains underexplored.In this study,a total of 77 trees were sampled from Northeast China to model the vertical distribution of branch diameter and length within each whorl along the crown.Several commonly used functions were taken as the alternative model forms,and the quantile regression method was employed and compared with the classical two-step modeling approach.The analysis incorporated stand,tree,and competition factors,with a particular focus on how these factors influence branches of varying sizes.The modified Weibull function was chosen as the optimal model,due to its excellent performance across all quantiles.Eight quantile regression curves(ranging from 0.20 to 0.85)were combined to predict branch diameter,while seven curves(ranging from 0.20 to 0.80)were used for branch length.The results showed that the quantile regression method outperformed the classical approach at model fitting and validation,likely due to its ability to estimate different rates of change across the entire branch size distribution.Lager branches in each whorl were more sensitive to changes in DBH,crown length(CL),crown ratio(CR)and dominant tree height(H_(dom)),while slenderness(HDR)more effectively influenced small and medium-sized branches.The effect of stand basal area(BAS)was relatively consistent across different branch sizes.The findings indicate that quantile regression is a good way not only a more accurate method for predicting branch size but also a valuable tool for understanding how branch growth responds to stand and tree factors.The models developed in this study are prepared to be further integrated into tree growth and yield simulation system,contributing to the assessment and promotion of wood quality.
基金supported by the National Natural Science Foundation of China (Project No.42375192)the China Meteorological Administration Climate Change Special Program (CMA-CCSP+1 种基金Project No.QBZ202315)support by the Vector Stiftung through the Young Investigator Group"Artificial Intelligence for Probabilistic Weather Forecasting."
文摘Despite the maturity of ensemble numerical weather prediction(NWP),the resulting forecasts are still,more often than not,under-dispersed.As such,forecast calibration tools have become popular.Among those tools,quantile regression(QR)is highly competitive in terms of both flexibility and predictive performance.Nevertheless,a long-standing problem of QR is quantile crossing,which greatly limits the interpretability of QR-calibrated forecasts.On this point,this study proposes a non-crossing quantile regression neural network(NCQRNN),for calibrating ensemble NWP forecasts into a set of reliable quantile forecasts without crossing.The overarching design principle of NCQRNN is to add on top of the conventional QRNN structure another hidden layer,which imposes a non-decreasing mapping between the combined output from nodes of the last hidden layer to the nodes of the output layer,through a triangular weight matrix with positive entries.The empirical part of the work considers a solar irradiance case study,in which four years of ensemble irradiance forecasts at seven locations,issued by the European Centre for Medium-Range Weather Forecasts,are calibrated via NCQRNN,as well as via an eclectic mix of benchmarking models,ranging from the naïve climatology to the state-of-the-art deep-learning and other non-crossing models.Formal and stringent forecast verification suggests that the forecasts post-processed via NCQRNN attain the maximum sharpness subject to calibration,amongst all competitors.Furthermore,the proposed conception to resolve quantile crossing is remarkably simple yet general,and thus has broad applicability as it can be integrated with many shallow-and deep-learning-based neural networks.
基金supported by the National Natural Science Foundation of China(Grant Nos.52109144,52025094 and 52222905).
文摘This paper introduces a novel approach for parameter sensitivity evaluation and efficient slope reliability analysis based on quantile-based first-order second-moment method(QFOSM).The core principles of the QFOSM are elucidated geometrically from the perspective of expanding ellipsoids.Based on this geometric interpretation,the QFOSM is further extended to estimate sensitivity indices and assess the significance of various uncertain parameters involved in the slope system.The proposed method has the advantage of computational simplicity,akin to the conventional first-order second-moment method(FOSM),while providing estimation accuracy close to that of the first-order reliability method(FORM).Its performance is demonstrated with a numerical example and three slope examples.The results show that the proposed method can efficiently estimate the slope reliability and simultaneously evaluate the sensitivity of the uncertain parameters.The proposed method does not involve complex optimization or iteration required by the FORM.It can provide a valuable complement to the existing approximate reliability analysis methods,offering rapid sensitivity evaluation and slope reliability analysis.
基金supported by the National Natural Science Foundation of China(Grant No.52109010)the Postdoctoral Science Foundation of China(Grant No.2021M701047)the China National Postdoctoral Program for Innovative Talents(Grant No.BX20200113).
文摘Copula functions have been widely used in stochastic simulation and prediction of streamflow.However,existing models are usually limited to single two-dimensional or three-dimensional copulas with the same bivariate block for all months.To address this limitation,this study developed a mixed D-vine copula-based conditional quantile model that can capture temporal correlations.This model can generate streamflow by selecting different historical streamflow variables as the conditions for different months and by exploiting the conditional quantile functions of streamflows in different months with mixed D-vine copulas.The up-to-down sequential method,which couples the maximum weight approach with the Akaike information criteria and the maximum likelihood approach,was used to determine the structures of multivariate Dvine copulas.The developed model was used in a case study to synthesize the monthly streamflow at the Tangnaihai hydrological station,the inflow control station of the Longyangxia Reservoir in the Yellow River Basin.The results showed that the developed model outperformed the commonly used bivariate copula model in terms of the performance in simulating the seasonality and interannual variability of streamflow.This model provides useful information for water-related natural hazard risk assessment and integrated water resources management and utilization.
基金the National Natural Science Foundation of China(NSFC)under No.52308473the National KeyR&DProgram under No.2022YFB2603301the China Postdoctoral Science Foundation funded project(Certificate Number:2023M743895).
文摘Purpose–This study aimed to facilitate a rapid evaluation of track service status and vehicle ride comfort based on car body acceleration.Consequently,a low-cost,data-driven approach was proposed for analyzing speed-related acceleration limits in metro systems.Design/methodology/approach–A portable sensing terminal was developed to realize easy and efficient detection of car body acceleration.Further,field measurements were performed on a 51.95-km metro line.Data from 272 metro sections were tested as a case study,and a quantile regression method was proposed to fit the control limits of the car body acceleration at different speeds using the measured data.Findings–First,the frequency statistics of the measured data in the speed-acceleration dimension indicated that the car body acceleration was primarily concentrated within the constant speed stage,particularly at speeds of 15.4,18.3,and 20.9 m/s.Second,resampling was performed according to the probability density distribution of car body acceleration for different speed domains to achieve data balance.Finally,combined with the traditional linear relationship between speed and acceleration,the statistical relationships between the speed and car body acceleration under different quantiles were determined.We concluded the lateral/vertical quantiles of 0.8989/0.9895,0.9942/0.997,and 0.9998/0.993 as being excellent,good,and qualified control limits,respectively,for the lateral and vertical acceleration of the car body.In addition,regression lines for the speedrelated acceleration limits at other quantiles(0.5,0.75,2s,and 3s)were obtained.Originality/value–The proposed method is expected to serve as a reference for further studies on speedrelated acceleration limits in rail transit systems.
基金Outstanding Youth Foundation of Hunan Provincial Department of Education(Grant No.22B0911)。
文摘In this paper,we introduce the censored composite conditional quantile coefficient(cC-CQC)to rank the relative importance of each predictor in high-dimensional censored regression.The cCCQC takes advantage of all useful information across quantiles and can detect nonlinear effects including interactions and heterogeneity,effectively.Furthermore,the proposed screening method based on cCCQC is robust to the existence of outliers and enjoys the sure screening property.Simulation results demonstrate that the proposed method performs competitively on survival datasets of high-dimensional predictors,particularly when the variables are highly correlated.
文摘This paper aims to explore the application of Extreme Value Theory (EVT) in estimating the conditional extreme quantile for time-to-event outcomes by examining the functional relationship between ambulatory blood pressure trajectories and clinical outcomes in stroke patients. The study utilizes EVT to analyze the functional connection between ambulatory blood pressure trajectories and clinical outcomes in a sample of 297 stroke patients. The 24-hour ambulatory blood pressure measurement curves for every 15 minutes are considered, acknowledging a censored rate of 40%. The findings reveal that the sample mean excess function exhibits a positive gradient above a specific threshold, confirming the heavy-tailed distribution of data in stroke patients with a positive extreme value index. Consequently, the estimated conditional extreme quantile indicates that stroke patients with higher blood pressure measurements face an elevated risk of recurrent stroke occurrence at an early stage. This research contributes to the understanding of the relationship between ambulatory blood pressure and recurrent stroke, providing valuable insights for clinical considerations and potential interventions in stroke management.
文摘As extremely important methods, Lp regression methods have attracted the attention of either theoretical or empirical researchers all over the world. As special cases of that, quantile and expectile regression (with p = 1 and p = 2, respectively) are well acquainted with. In contrast with them, general Lp regression (with p equals 1 and 2) has recently been found to have many unexpected properties by some studies, especially when 1 p Lp quantile regression under various p settings and shows some recent advances in Lp quantile regression, theoretically and empirically.
基金supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea(NRF-2022S1A5A2A01038422).
文摘In the context of the rapidly growing demand for green investments and the need to combat climate change,this study contributes to the emerging literature on green investments by exploring the time-frequency connectedness between green bonds(GBs)and green equities.Specifically,we examine the degree of connection between GBs and green equities,the extent to which these markets influence each other,and which one is the primary net transmitter versus the net receiver of shocks under diverse market conditions.To accomplish these objectives,we use the wavelet-based Quantile-on-Quantile(QQ),dynamic conditional correlation(DCC),portfolio implications,and Quantile VAR approaches.The results show that GBs and green equities have a strong positive connection,depending on time and frequency domains.However,a negative association between GBs and green equities is observed during periods of crisis,highlighting GBs’ability to hedge green equity portfolios.The portfolio strategies demonstrate that investors require to invest in the Green Economy equity and S&P GB portfolio to reach the highest level of hedging effectiveness.The findings further imply that the Global Water Equity Index transmits the highest spillover to other green assets,while the Green Economy Equity Index receives the most spillover from other assets.The pairwise volatility connectivity reveals that most pairs have minimal quantile dependence,indicating the potential for diversification across the GB and green equity pairs.These findings have significant implications for investors and policymakers concerned with green investments and climate change mitigation.
基金supported by the Jiangxi Provincial Natural Science Foundation(20212ACB211003)the National Natural Science Foundation of China(No.71671029).
文摘Value at risk(VaR)and expected shortfall(ES)have emerged as standard measures for detecting the market risk of financial assets and play essential roles in investment decisions,external regulations,and risk capital allocation.However,existing VaR estimation approaches fail to accurately reflect downside risks,and the ES estimation technique is quite limited owing to its challenging implementation.This causes financial institutions to overestimate or underestimate investment risk and finally leads to the inefficient allocation of financial resources.The main purpose of this study is to use machine learning to improve the accuracy of VaR estimation and provide an effective tool for ES estimation.Specifically,this study proposes a VaR estimator by combining quantile regression with“Mogrifier”recurrent neural networks to capture the“long memory”and“clustering”properties of financial assets;while for estimating ES,this study directly models the quantile of assets and employs generative adversarial networks to generate future tail risk scenarios.In addition to the typical properties of financial assets,the model design is also consistent with heterogeneous market theory.An empirical application to four major global stock indices shows that our model is superior to other existing models.
文摘There is a substantial body of empirical research that has found the fund return distributions to exhibit pronounced peakiness, heavy tails, and skewness, deviating from a normal distribution. Addressing the limitations of the traditional Sharpe ratio, which assumes a normal distribution of returns and uses standard deviation to measure investment risk, this paper primarily employs the Value at Risk (VaR) based on Lp quantile to adjust excess returns of funds. This method offers superior robustness, is capable of capturing asymmetry and heavy-tailed characteristics, and is more flexible, providing a better description of the tail risk in fund returns. Empirical studies have shown that using the Sharpe ratio corrected with the Lp quantile is feasible for evaluating and ranking the performance of open-end funds.
文摘Analyzing the interdependencies among financial institutions is critical for designing systemic risk monitoring mechanisms;however,most existing research focuses on the first moment of the return distribution,which falls into the conventional models of choice under risk.Previous literature has observed the scarcity of investors’attention and processing power,which makes the traditional theory of choice under risk more vulnerable and brings the salience theory that accommodates investors’cognitive limitations to our attention.Motivated by evidence of salience theory value(STV)containing unique information not captured by traditional higher-order moments,we employ a quantile connectedness approach to examine the STV interconnectedness of China’s systemically important banks(C-SIBs).The quantile approach allows us to uncover the dynamic STV interconnectedness of C-SIBs under normal,bearish,and bullish market conditions and is well-suited to extreme risk problems.Our results show that the C-SIBs system is asymmetrically interconnected across quantiles and at higher levels under bullish than bearish market conditions.Principally,a bank’s performance in the C-SIBs system depends on its systemic importance and market conditions.Furthermore,the comparative analysis indicates that STV could provide more information than higher-order moments in capturing the dynamic change in the C-SIBs system and detecting some market events more precisely.These results have important implications for policymakers and market participants to formulate regulatory policy and design risk management strategies.
基金the financial support from Science and Engineering Research Board,Department of Science&Technology,Government of India,under the scheme Early Career Research Award(file no.ECR/2017/002416)Dr.Sharma also acknowledges Banaras Hindu University,Varanasi,India,for providing financial support as seed grant under the Institute of Eminence Scheme(Scheme no.Dev.6031).
文摘This article introduces a three-parameter Lehman-type t distribution having 2 degrees of freedom,that is capable of fitting positive and negative skewed data sets.It is shown that the density and hazard functions of the proposed distribution are uni-model.Ordinary moments,entropy measure,ordering,identifiability and order statistics are investigated.Since the quantile function is explicitly defined,quantile-based statistics are also discussed for the proposed distribution.These properties include measures of skewness and kurtosis,L-moments,quantile density and hazard functions,mean residual life function and Parzen's score function.Mechanisms of maximum likelihood,bias correction and matching of percentiles are employed for estimating the unknown parameters of the distribution.Simulation experiments are conducted to compare the performance of these three estimation methods.A real-life data set consisting of strength of glass fibres is fitted to show the adequacy of the proposed distribution over some extensions of the normal and t distributions.Parametric regression model is developed along with its parameter estimation using the maximum likelihood approach.Simulation study for the regression model is also presented that endorsed the asymptotic properties of the estimators.
文摘Since its inception,the epsilon distribution has piqued the interest of statisticians.It has been successfully used to solve a variety of statistical problems.In this article,we propose to use the quadratic rank transmutation map mechanism to extend this distribution.This mechanism is not new;it was already used to improve the modeling capabilities of a number of existing distributions.For the original epsilon distribution,we expect the same benefits.As a result,we implement the transmuted epsilon distribution as a flexible three-parameter distribution with a bounded domain.We demonstrate its key features,focusing on the properties of its distributional mechanism and conducting quantile and moment analyses.Applications of the model are presented using two data sets.We also perform a regression analysis based on this distribution.
基金National Key Research and Development Plan of China,No.2023YFF0805703National Natural Science Foundation of China,No.42271268。
文摘Heatwaves are becoming increasingly frequent and severe,posing escalating risks to ecosystems and human well-being.While soil moisture(SM)deficits are recognized as important contributors to heatwave amplification,their spatially heterogeneous impacts across the Northern Hemisphere remain insufficiently understood.In this study,we analyze ERA5 reanalysis data(1980-2022)to investigate trends in heatwave frequency,intensity,and duration,as well as their sensitivity to SM variability.Our results show robust increases in heatwave occurrence(0.76 events per decade),intensity(0.81℃per decade),and average duration(0.40 days per decade),with extreme events,as represented by maximum intensity and duration,rising at even faster rates(2.18℃per decade and 0.83 days per decade,respectively).Strong negative correlations are observed between SM deficits and heatwave metrics,with the magnitude of this relationship varying across land cover types and heatwave severity levels.Quantile regression reveals that SM reductions have a greater impact at higher quantiles for most indicators.Cropland exhibits the highest sensitivity to SM anomalies,whereas forests show more resilience due to their superior water retention capacities.These findings underscore the crucial role of land-atmosphere interactions in shaping heatwave extremes,providing a scientific basis for enhancing early warning and adaptation strategies in the context of ongoing climate change.
文摘As the world’s fourth most populous country,Indonesia presents challenges and opportunities for sustainable energy progress,offering a critical context to investigate green human development(GHD).This study uniquely contributes to the literature by employing the planetary pressures-adjusted human development index(PHDI)as an indicator of GHD,which integrates environmental impacts into human development.Using static and dynamic econometric methods,including the quantile regression and autoregressive distributed lag model,it explores the impacts of renewable and nonrenewable energy consumption on GHD.The findings demonstrate that renewable energy currently has a detrimental impact on GHD due to its limited adoption and high costs.Conversely,nonrenewable energy positively influences GHD,as it is the primary energy source in the country and is becoming more efficient at reducing emissions.However,the study finds that greater use of renewable energy reduces its adverse effects,suggesting that as renewable energy technologies become more cost-effective and widely implemented,their initial adverse effects could be mitigated,leading to improved long-term GHD outcomes.These findings carry important implications for Indonesia,where the govern‐ment is striving to expand renewable energy capacity while promoting equitable development across its archi‐pelagic regions.They underscore the critical role of energy policy in balancing economic,social,and environmental goals,contributing meaningfully to the country’s sustainable development agenda.
文摘Accurate forecasting of oil production is essential for optimizing resource management and minimizing operational risks in the energy sector. Traditional time-series forecasting techniques, despite their widespread application, often encounter difficulties in handling the complexities of oil production data, which is characterized by non-linear patterns, skewed distributions, and the presence of outliers. To overcome these limitations, deep learning methods have emerged as more robust alternatives. However, while deep neural networks offer improved accuracy, they demand substantial amounts of data for effective training. Conversely, shallow networks with fewer layers lack the capacity to model complex data distributions adequately. To address these challenges, this study introduces a novel hybrid model called Transfer LSTM to GRU (TLTG), which combines the strengths of deep and shallow networks using transfer learning. The TLTG model integrates Long Short-Term Memory (LSTM) networks and Gated Recurrent Units (GRU) to enhance predictive accuracy while maintaining computational efficiency. Gaussian transformation is applied to the input data to reduce outliers and skewness, creating a more normal-like distribution. The proposed approach is validated on datasets from various wells in the Tahe oil field, China. Experimental results highlight the superior performance of the TLTG model, achieving 100% accuracy and faster prediction times (200 s) compared to eight other approaches, demonstrating its effectiveness and efficiency.
