427 papers
This paper proposes and validates kernel smoothing and maximum likelihood methods for estimating parameters of the Vasicek process under microstructure noise, demonstrating their strong consistency and accuracy through theoretical analysis, numerical simulations, and empirical application to Shanghai Stock Exchange data.
This study proposes a hybrid GAM-LightGBM framework with selective conformal calibration and a nesting procedure to enhance the accuracy and coherence of probabilistic load forecasts across multiple European countries under varying weather uncertainties.
This study validates a seven-attribute Q-matrix for assessing clinical reasoning in a nationwide Korean medical examination using a DINA model, demonstrating strong known-groups validity through significant mastery differences between third- and fourth-year students.
This paper presents a hierarchical Bayesian spatial modeling framework that generates probabilistic, neighborhood-specific heat-health warnings across California by estimating varying temperature thresholds linked to emergency department visits, thereby enabling equitable resource allocation and forming the basis for the state's official CalHeatScore system.
This study demonstrates that the precision-weighting coefficient (γ) in the CPSP-8 model is a pipeline-sensitive, bounded scalar that fails to generalize as a portable biomarker across tasks unless precision gain is sufficiently large, whereas other decision parameters function merely as mode switches.
This paper proposes a method to derive calibrated, two-sided confidence intervals for the true Area Under the Curve (AUC) in Positive-Unlabeled learning by exactly recovering the target AUC from observable metrics and propagating the uncertainty of the estimated positive fraction, thereby addressing the bias and lack of reliability in current performance evaluations.
This study evaluates classical, machine learning, and deep learning models for forecasting weekly dengue cases in Marikina City using a leakage-safe temporal design, finding that while simple baselines performed well during internal validation, the LSTM model achieved superior predictive accuracy on the 2025 out-of-time holdout despite a general tendency to underpredict outbreak-level weeks.
This study employs a participant-specific, uncertainty-aware Bayesian hierarchical model to analyze a within-participant experiment, revealing that AI-generated voices paired with AI avatars yield the most favorable engagement and cognitive load profile while highlighting significant individual heterogeneity in responses to audiovisual combinations.
This study of Library and Information Science publications reveals that the open access citation advantage is not uniform but is instead topic-dependent and largely driven by health-sciences-adjacent subfields, while being significantly attenuated when controlling for journal-level citation baselines.
This paper introduces flexible linear, quadratic, and second-order component-position screening models that utilize position zero to encode component omission, enabling efficient estimation of nonlinear absolute-position effects and the construction of optimal designs for order-of-addition experiments where only a subset of components is available per run.