Search Results (62)

Huber regression serves as a prominent robust alternative to ordinary least squares (OLS), particularly in the presence of heavy-tailed error distributions. While the asymptotic relative efficiency (ARE) of Huber regression is well documented for the standard normal distribution, its worst-case efficiency across the class of all continuous and symmetric error distributions remains an important theoretical question. In this paper, we establish positive lower bounds for the ARE of Huber regression relative to OLS. By strategically selecting the robustification parameter based on the moments or quantiles of the error distribution, we first prove that the ARE is uniformly bounded away from zero across all continuous and symmetric error distributions. This result guarantees a baseline level of efficiency for Huber regression, sharing a similar theoretical spirit with the celebrated lower bound of the Wilcoxon rank estimator. Utilizing the empirical process theory, we further establish that the relative efficiency of Huber regression remains unchanged if the theoretical tuning parameter is replaced by an estimator with a suitable convergence rate. Simulation studies are conducted to examine the performance of Huber regression under the proposed tuning strategies. Full article

Increasing the carbon sink function of forests is critical for achieving carbon (C) neutrality in the context of global climate change. Past studies have focused on the estimation of forest biomass or C storage, while those on forest C sink potential remain limited. In particular, there remain few systematic investigations to define the forest C sink, to characterize the synergistic influencing factors, and to develop related quantitative analysis methods. The development of scientific C enhancement strategies requires the construction of C density-age models integrating multiple stand factors. These models allow accurate quantification of the gap (∆C) between actual and maximum C sequestration capacity. This study used permanent sample plot data to develop and validate a novel multi-model assessment approach for quantifying the C sink potential of Larix olgensis plantations in Heilongjiang Province, China, and to translate the results into precise management tools. An Average-Level Model (ALM) was established to define baseline C sequestration. Three innovative potential assessment models were then proposed: (1) the Empirical Upper Boundary Model (PLM1); (2) the Dummy Variable Model (PLM2); and (3) the Quantile Regression Model (PLM3). These models define the maximum C sequestration capacity from distinct perspectives. PLM1 (R² = 0.7910) characterized the theoretical upper limit of C sink potential (79.86 Mg·ha⁻¹), making it suitable for macro-strategic goal setting, though it is somewhat dependent on extreme data points. PLM2 (R² = 0.7943) achieved the best fit, and when combined with measurable stand conditions (site class index [SCI] > 16 m, stand density index [SDI] > 800 trees·ha⁻¹), it provides clear guidance for management practices. Although PLM3 showed a lower goodness-of-fit (R² = 0.1056), it provided reasonable parameter estimates and robust predictions, offering a reliable upper-bound reference for C sink project planning and risk control. At a stand age of 60 years (yr), the C sink enhancement potentials (“∆” C) corresponding to the three models were 15.73, 14.48, and 13.26 Mg·ha⁻¹, representing increases of 24.53%, 22.58%, and 20.68%, respectively, over the average level (64.13 Mg·ha⁻¹); the peak C sequestration rates of the models were 104.3%, 82.7%, and 60.5% higher than that of the ALM, with peak times occurring earlier at 9, 7, and 11 yr, respectively, underscoring the importance of the early management. The multi-model assessment approach developed here facilitates “precision carbon enhancement” by quantifying C sink potential across its theoretical, achievable, and robust upper-bound dimensions. This quantification provides both mechanistic insights into C sequestration processes and a critical link between theoretical understanding and practical forest management. This work holds significant value for advancing forestry C sinks in service of national strategies. Full article

Medium-term 5G/NB-IoT planning is made difficult by simultaneous uncertainty in device adoption and per-device traffic behavior because deterministic point forecasts do not quantify overload risk or support reliability-based capacity decisions. A diffusion-to-dimensioning workflow is proposed in which S-curve adoption modeling, bounded usage priors, scenario stress testing, and Monte Carlo uncertainty propagation are combined to generate predictive demand distributions, exceedance curves, and quantile-based capacity rules. The framework is applied to a Great Britain case study for 2025–2029 using smart meter deployment data and an M2M-based proxy for asset-tracking adoption. Analysis shows that planning-year upper-tail outcomes are driven primarily by asset-tracking usage uncertainty rather than by proxy scale alone. A ±30% perturbation of the AT adoption anchor changes $Q_{0.95}$ by approximately ±29.8%, whereas stressed AT usage increases $Q_{0.95}$ by 74.4%. Plausible positive dependence among key AT operational inputs further raises $Q_{0.95}$ by 18.3–22.5%. Limited hold-out evaluation provides strong out-of-sample support for the smart meter adoption stage and plausibility-only support for the shorter AT proxy. The framework is intended for medium-term, data-lean planning settings and is designed to support transparent risk-based capacity decisions rather than deterministic point sizing. Full article

We introduce the one-parameter bounded p-exponential distribution on $(0, p+1)$, which includes the uniform model as a special case and converges pointwise to the exponential law as $p\to \infty$. Closed-form expressions are derived for the CDF and PDF, the survival function, an explicit increasing-failure-rate hazard function, the quantile function (enabling inversion-based simulation), moments, and entropy, along with a constructive scaled beta or Kumaraswamy representation. We also establish stochastic ordering with respect to p in stop-loss and increasing convex order, formalizing how dispersion varies with the parameter while preserving the mean scale. Inference is discussed under parameter-dependent support, a non-regular setting, and we develop and compare several estimation procedures, including a likelihood-based boundary MLE, a variance-matching method-of-moments estimator, and Bayesian estimation under a gamma prior implemented via numerical quadrature or MCMC. Monte Carlo simulation studies evaluate finite-sample performance and interval behavior, and two real-world applications in survival and reliability analysis illustrate competitive goodness-of-fit relative to standard benchmark models. Full article

High-rise elevator group control systems operate under pronounced nonstationarity during commuting peaks, post-event surges, and capacity degradation, where the waiting time distribution becomes right-tail heavy and stresses service-level agreements (SLAs) defined by coverage and high-quantile targets. At the same time, the time-of-use tariffs and carbon constraints sharpen the tension between peak-power control, energy savings, and service capacity. This paper proposes a two-layer resilience scheduling framework that integrates queueing-based planning with safe reinforcement learning (RL) fine-tuning. In the planning layer, parsimonious queueing approximations and scenario-based evaluation construct a finite set of implementable mode cards and emergency switching cards; Sample Average Approximation (SAA) combined with Conditional Value-at-Risk (CVaR) constraints filter candidates to enforce tail-risk-aware service limits while keeping power demand within a prescribed envelope. In the execution layer, online dispatch is formulated as a constrained Markov decision process; within the planning layer limits, action masking and Lagrangian safe RL learn small adaptive adjustments to suppress tail-waiting risk and improve recovery dynamics without increasing peak-power commitments. The experiments under morning peaks and post-event surges confirm tail risk reduction and accelerated recovery. For partial outages, the framework prioritizes SLA coverage and recovery speed, accepting a bounded increase in tail risk as a manageable trade-off. Throughout all tests, peak power remains within the prescribed limits. Improvements persist across random seeds and demand fluctuations, indicating distributional robustness and cross-scenario generalization. Ablation studies further reveal complementary roles: removing the planning layer CVaR screening worsens tail performance, while removing the execution layer action masking increases constraint violations and destabilizes recovery. Full article

Rapid urbanization, environmental degradation and climate variability are intensifying the exposure of urban populations to multiple, interacting hazards in megacities. In India’s capital, Delhi, extreme heat, worsening air quality and flood-related stress overlap in impacted areas, exacerbated by high population density in low-lying zones and extensive built-up cover. This study develops an integrated spatial framework for assessing relative multi-hazard risk potential in Delhi by combining remote sensing, climate reanalysis, land use and demographic datasets into a predictive modeling system to support urban resilience planning. A comprehensive suite of twenty-two predictors representing thermal stress, air quality, surface indices, topography, hydrology, land use land cover (LULC), and demographic data was derived from diverse Earth observation sources. A cloud-native workflow leveraging Google Earth Engine (GEE) and Python 3 harmonized these predictors to train a Light Gradient Boosting Machine (LightGBM) model with five-fold spatial cross-validation. Quantile regression was used to estimate lower (P10) and upper (P90) predictive bounds, which are interpreted here as empirical predictive intervals around the modeled risk surface rather than as a strict separation of different uncertainty types, while SHapley Additive exPlanations (SHAP) decomposed the non-linear contributions of individual features. The model achieved predictive accuracy (R² = 0.98, MAE = 0.01), with residuals centered near zero and consistent performance across spatial folds, demonstrating strong generalizability. Road density (63.4%) and population density (25.9%) emerged as the primary predictors of the modeled risk surface, followed by building density and NO₂ concentration. Conversely, vegetation cover (NDVI) functioned as a critical mitigating buffer. Spatial risk maps identified persistent high-risk clusters in eastern and northeastern Delhi, coinciding with dense transport networks and industrial zones. The integrated P90 mapping framework provides spatially explicit and uncertainty-aware information on relative multi-hazard risk potential to guide targeted interventions, such as transport corridor mitigation and urban greening in Delhi and other rapidly urbanizing cities. Full article

Inertial localization is particularly valuable in GPS-denied environments such as indoors. However, localization using only Inertial Measurement Units (IMUs) suffers from drift caused by motion-process noise and sensor biases. This paper introduces Uncertainty-aware Map-constrained Inertial Localization (UMLoc), an end-to-end framework that jointly models IMU uncertainty and map constraints to achieve drift-resilient positioning. UMLoc integrates two coupled modules: (1) a Long Short-Term Memory (LSTM) quantile regressor, which estimates the specific quantiles needed to define $68\%$, $90\%$ and $95\%$ prediction intervals serving as a measure of localization uncertainty and (2) a Conditioned Generative Adversarial Network (CGAN) with cross-attention that fuses IMU dynamic data with distance-based floor-plan maps to generate geometrically feasible trajectories. The modules are trained jointly, allowing uncertainty estimates to propagate through the CGAN during trajectory generation. UMLoc was evaluated on three datasets, including a newly collected 2-h indoor benchmark with time-aligned IMU data, ground-truth poses and floor-plan maps. Results show that the method achieves a mean drift ratio of $5.9\%$ over a $70\mathrm{m}$ travel distance and an average Absolute Trajectory Error (ATE) of $1.36\mathrm{m}$, while maintaining calibrated prediction bounds. Full article

Credit default prediction from firm-level accounting statements is central to risk management, yet the underlying financial data are highly sensitive and often siloed across banks, auditors, and platforms. Federated learning (FL) offers a practical route to collaborative modeling without centralizing raw records, but standard FL optimization can converge slowly under severe client heterogeneity, heavy-tailed accounting features, and label imbalance typical of default events. This paper proposes FedQuAD, a novel fast-converging FL algorithm that couples (i) quasi-Newton curvature aggregation on the server with a lightweight limited-memory update to accelerate global progress, (ii) a proximal variance-reduced local solver that stabilizes client drift under non-IID accounting distributions, and (iii) federated robust standardization of tabular financial ratios via secure aggregated quantile statistics to mitigate scale instability and outliers. FedQuAD is communication-efficient by design: It transmits compact gradient and curvature sketches and adapts local computation to each client’s stochasticity and drift. We provide convergence guarantees for strongly convex default-risk objectives (logistic and calibrated GLM losses) under bounded heterogeneity, and extend the analysis to nonconvex deep tabular models via expected stationarity bounds. Experiments on public credit-risk benchmarks with simulated cross-silo (institutional) partitions demonstrate that FedQuAD reaches target AUC and calibration error with substantially fewer communication rounds than representative baselines while maintaining privacy constraints compatible with secure aggregation and optional client-level differential privacy accounting. Full article

This paper presents a new continuous data model, the Unit Arcsine–Exponential distribution (UASED), a flexible data model on the unit interval. It is built up by an exponential-based arcsine-type transformation to allow it to represent a very wide range of shapes that can be used to model proportions and rates. A number of basic properties are obtained, such as closed-form formulas of the quantile function, moments, and entropy measures. Maximum likelihood and maximum product of spacings methods are developed to estimate parameters, and their performance is determined by Monte Carlo simulation, which shows that these methods can reasonably estimate the parameters and be stable over a variety of different parameter settings. To demonstrate that a model is practically useful, an application to real-world data on the reliability of devices in terms of failure time is discussed. The findings indicate that the UASED is a good fit to the data, in the sense that it is effective in terms of skewness and tail behavior and compares well or competes favorably with current unit distributions. All in all, the suggested model is a sparse alternative to model bounded data with sound inferential characteristics and high practical utility. Full article

Under non-independent and identically distributed (Non-IID) conditions, significant variations exist in local model updates across clients and training phases during the collaborative modeling process of differential privacy federated learning (DP-FL). Fixed clipping thresholds and noise scales struggle to accommodate these diverse update differences, leading to mismatches between local update intensity and noise perturbations. This imbalance results in data privacy leaks and suboptimal model accuracy. To address this, we propose a differential privacy federated learning method based on adaptive clipping thresholds. During each communication round, the server adaptively estimates the global clipping threshold for that round using a quantile strategy based on the statistical distribution of client update norms. Simultaneously, clients adaptively adjust their noise scales according to the clipping threshold magnitude, enabling dynamic matching of clipping intensity and noise perturbation across training phases and clients. The novelty of this work lies in a quantile-driven, round-wise global clipping adaptation that synchronizes sensitivity bounding and noise calibration across heterogeneous clients, enabling improved privacy–utility behavior under a fixed privacy accountant. Using experimental results on the rail damage datasets, our proposed method slightly reduces the attacker’s MIA ROC-AUC by 0.0033 and 0.0080 compared with Fed-DPA and DP-FedAvg, respectively, indicating stronger privacy protection, while improving average accuracy by 1.55% and 3.35% and achieving faster, more stable convergence. We further validate its effectiveness on CIFAR-10 under non-IID partitions. Full article

Although particulate matter has been associated with sleep problems, the effects of PM_2.5-bound organophosphate esters (OPEs) on children’s sleep remain unclear. OPEs have neurotoxic and endocrine-disrupting effects that may disrupt sleep–wake regulation during neurodevelopment, supporting biological plausibility for sleep impacts. In this study, we quantified the individual and mixture effects of PM_2.5-bound OPEs on the sleep disorder domain. This cross-sectional study included 110,169 children aged 6–18 years from primary and secondary schools in the Pearl River Delta (PRD), China. Sleep disorders were evaluated using the validated Sleep Disturbance Scale for Children (SDSC). Elastic net and mixed effect models identified specific OPE–sleep associations, while weighted quantile sum regression evaluated mixture effects. All odds ratios indicate a change in the likelihood of sleep disorders per interquartile range (IQR) increase in OPE concentrations. The strongest individual associations were observed for TDCIPP with short sleep duration (OR = 1.56–1.61; moderate association), TEHP with short sleep duration (OR = 1.59–1.64; moderate association), and TPHP with overall sleep disorder (OR = 1.32–1.42; modest association). Combined OPE exposure was positively associated with all sleep disorder domains (ORs = 2.02–2.85; moderate-to-large associations). These results indicate that inhaling PM_2.5-bound OPE mixtures could negatively impact children’s sleep health. This emphasizes a critical developmental period and highlights the importance of public health concerns related to emerging airborne contaminants. Full article

Reliable probabilistic wind speed forecasts are essential for integrating renewable energy into power grids and managing operational uncertainty. This study compares Quantile Regression Forests (QRF), Bayesian Additive Regression Trees (BART), and Gaussian Process Regression (GPR) under at-site and regional pooled frameworks using 21 years (2000–2020) of daily wind data from eleven stations in New South Wales and Queensland, Australia. Models are evaluated via strict year-based holdout validation across seven metrics: RMSE, MAE, R², bias, correlation, coverage, and Continuous Ranked Probability Score (CRPS). Regional QRF achieves exceptional point forecast stability with minimal RMSE increase but suffers persistent under-coverage, rendering probabilistic bounds unreliable. BART attains near-nominal coverage at individual sites but experiences catastrophic calibration collapse under regional pooling, driven by fixed noise priors inadequate for spatially heterogeneous data. In contrast, GPR maintains robust probabilistic skill regionally despite larger point forecast RMSE penalties, achieving the lowest overall CRPS and near-nominal coverage through kernel-based variance inflation. Variable importance analysis identifies surface pressure and minimum temperature as dominant predictors (60–80%), with spatial covariates critical for regional differentiation. Operationally, regional QRF is prioritised for point accuracy, regional GPR for calibrated probabilistic forecasts in risk-sensitive applications, and at-site BART when local data suffice. These findings show that Bayesian machine learning methods can effectively navigate the trade-off between local specificity and regional pooling, a challenge common to wind forecasting in diverse terrain globally. The methodology and insights are transferable to other heterogeneous regions, providing guidance for probabilistic wind forecasting and renewable energy grid integration. Full article

The estimation of correlation matrices in high-dimensional data streams presents a fundamental conflict between computational efficiency and statistical robustness. Moment-based estimators, such as Pearson’s correlation, offer linear $O\left(N\right)$ complexity but lack robustness. In contrast, high-breakdown methods like the minimum covariance determinant (MCD) are computationally prohibitive ($O(N{p}^2+p^3)$) for real-time applications. This paper introduces Fast Outlier-Robust Correlation Estimation (FORCE), a streaming algorithm that performs adaptive coordinate-wise trimming using the $P^2$ algorithm for streaming quantile approximation, requiring only $O\left(p\right)$ memory independent of stream length. We evaluate FORCE against six baseline algorithms—including exact trimmed methods (TP-Exact, TP-TER) that use $O(NlogN)$ sorting with $O\left(Np\right)$ storage—across five benchmark datasets spanning synthetic, financial, medical, and genomic domains. FORCE achieves speedups of approximately 470× over FastMCD and 3.9× over Spearman’s rank correlation. On S&P 500 financial data, coordinate-wise trimmed methods substantially outperform FastMCD: TP-Exact achieves the best RMSE (0.0902), followed by TP-TER (0.0909) and FORCE (0.1186), compared to FastMCD’s 0.1606. This result demonstrates that coordinate-wise trimming better accommodates volatility clustering in financial time series than multivariate outlier exclusion. FORCE achieves 76% of TP-Exact’s accuracy while requiring ${10}^4\times$ less memory, enabling robust estimation in true streaming environments where data cannot be retained for batch processing. We validate the 25% breakdown point shared by all IQR-based trimmed methods using the ODDS-satellite benchmark (31.7% contamination), confirming identical degradation for FORCE, TP-Exact, and TP-TER. For memory-constrained streaming applications with contamination below 25%, FORCE provides the only viable path to robust correlation estimation with bounded memory. Full article

A Hermite-based framework for reliability assessment within the limit state method is developed in this paper. Closed-form design quantiles under a four-moment Hermite density are derived by inserting the Gaussian design quantile into a calibrated cubic translation. Admissibility and implementation criteria are established, including a monotonicity bound, a positivity condition for the platykurtic branch, and a balanced Jacobian condition for the leptokurtic branch. Material data for the yield strength and ductility of structural steel are fitted using moment-matched Hermite models and validated through goodness-of-fit tests. A truss structure is subsequently analysed to quantify how non-Gaussian input geometry influences structural resistance and its associated design value. Variance-based Sobol sensitivity analysis shows that departures of the radius distribution toward negative skewness and higher kurtosis increase the first-order contribution of geometric variables and thicken the lower tail of the resistance distribution. The closed-form Hermite design resistances agree closely with numerical integration results and reveal systematic deviations from FORM estimates, which depend solely on the mean and standard deviation. Monte Carlo simulations confirm these trends and highlight the slow convergence of tail quantiles and higher-order moments. The proposed approach remains fully compatible in the Gaussian limit and offers a practical complement to EN 1990 verification procedures when skewness and kurtosis have a significant influence on design quantiles. Full article

This paper introduces the Marshall–Olkin unit-exponentiated-half-logistic (MO-UEHL) distribution, a novel three-parameter model designed to enhance the flexibility of the unit-exponentiated-half-logistic distribution through the incorporation of the Marshall–Olkin transformation. Defined on the unit interval $(0,1)$, the MO-UEHL distribution is well-suited for modeling proportional data exhibiting asymmetry. The Marshall–Olkin tilt parameter $\alpha$ explicitly controls the degree and direction of asymmetry, enabling the density to range from highly right-skewed to nearly symmetric unimodal forms, and even to left-skewed configurations for certain parameter values, thereby offering a direct mathematical representation of symmetry breaking in bounded proportional data. The resulting model achieves this versatility without relying on exponential terms or special functions, thus simplifying computational procedures. We derive its key mathematical properties, including the probability density function, cumulative distribution function, survival function, hazard rate function, quantile function, moments, and information-theoretic measures such as the Shannon and residual entropy. Parameter estimation is explored using maximum likelihood, maximum product spacing, ordinary and weighted least-squares, and Cramér–von Mises methods, with simulation studies evaluating their performance across varying sample sizes and parameter sets. The practical utility of the MO-UEHL distribution is demonstrated through applications to four real datasets from environmental and engineering contexts. The results highlight the MO-UEHL distribution’s potential as a valuable tool in reliability analysis, environmental modeling, and related fields. Full article