427 papers
This paper establishes a thermodynamic framework for data-driven modeling based on Jaynes' Maximum Entropy Principle, which treats inference as a physical process governed by a First Law of informational thermodynamics and utilizes heat capacity anomalies to endogenously identify optimal model architectures without ad-hoc regularization.
This paper introduces "geostatistical reachability," a simulation-calibrated framework that qualifies spatial model classes by assessing whether any admissible member can reproduce specific structural diagnostics before predictive ranking, thereby distinguishing structural compatibility and observability from mere predictive performance.
This paper introduces a new six-parameter generalized Weibull-Kumaraswamy-Lomax distribution using the T-X family framework, deriving its fundamental functions and proposing a novel linear series expansion for its probability density function to simplify the computation of raw moments and entropy.
This paper argues that the forecast skill of machine learning models for urban electricity demand is primarily determined by the intrinsic predictability of a city's historical load patterns (specifically seasonal variance and temperature sensitivity) rather than by the city's structural characteristics or geographic location.
This study establishes the Top-Down Scale (TDS) algorithm as a computationally efficient and statistically robust benchmark for Multiscale Geographically and Temporally Weighted Regression, demonstrating its superiority over existing methods in parameter recovery and uncertainty calibration while revealing distinct multiscale urban drivers in São Paulo through a longitudinal analysis of 34,938 census observations.
This study develops and validates a Stacking-based ensemble machine learning model using 5.7 million viral load records from Uganda to accurately predict the next visit viral load for ART clients, achieving superior performance with an R² of 0.85 and utilizing LIME for model interpretability.
This paper introduces a model-independent workflow to distinguish genuine local scale dependence from artificial trends caused by regional mixing in urban block-size distributions, demonstrating through a comprehensive analysis of Tokyo that while regional heterogeneity flattens the distribution tail, smaller blocks exhibit significant local steepening that cannot be explained by compositional effects alone.
This paper proposes and validates two types of uncertain single-index models using semiparametric least-squares estimation with kernel and B-spline methods to effectively handle complex, imprecise data, as demonstrated through simulation studies and a practical weather application.
This paper presents the theoretical development and structure of the Practitioner Allegiance Risk Scale (PARS-6), a new six-item instrument designed to systematically screen and document investigator allegiance-related characteristics during data extraction in systematic reviews of intervention research, while explicitly clarifying that it serves as a descriptive risk indicator rather than a definitive measure of bias.
This study utilizes SciVal bibliometric data and forecasting models to reveal that while Philippine State Universities have significantly increased publication volumes from 2015 to 2025, substantial and persistent regional disparities in research output and impact—led by Region III—will continue through 2030 without targeted capacity-building investments rather than uniform incentives.