Results 161 to 170 of about 204,570 (278)

Uncertainty Calibration in Molecular Machine Learning: Comparing Evidential and Ensemble Approaches

Chemistry – A European Journal, EarlyView.
ABSTRACT Machine learning (ML) models are increasingly used in quantum chemistry, but their reliability hinges on uncertainty quantification (UQ). In this study, we compare two prominent UQ paradigms—deep evidential regression (DER) and deep ensembles—on the QM9 and WS22 datasets, with a specific emphasis on the role of post hoc calibration.
Bidhan Chandra Garain, Max Pinheiro Jr., Matheus de Oliveira Bispo, Mario Barbatti   +3 more
wiley   +1 more source

Quantitative Charakterisierung der inhomogenen Anströmung der Filterschicht bei der Staubabscheidung mit Oberflächenfiltern unter Verwendung von Markerpartikeln – Teil 1: Das methodische Vorgehen Quantitative Characterization of the Inhomogeneous Inflow of the Filter Layer During Dust Collection with Surface Filters Using Marker Particles – Part 1: Methodical Approach

Chemie Ingenieur Technik, EarlyView.
Die inhomogene Anströmung der Filterschicht bei der Staubabscheidung mit Oberflächenfiltern wird mit einem neuen methodischen Vorgehen unter Verwendung von Markerpartikeln am Beispiel eines ausgewählten Nadelfilz‐Filtermediums experimentell untersucht.
Felix Belter, Qian Zhang
wiley   +1 more source

A scalar-on-quantile-function approach for estimating short-term health effects of environmental exposures. [PDF]

Biometrics
Zhang Y, Chang HH, Warren JL, Ebelt ST.
europepmc   +1 more source

An Investigation of Quantile Function Estimators Relative to Quantile Confidence Interval Coverage. [PDF]

Commun Stat Theory Methods, 2015
Wei L, Wang D, Hutson AD.
europepmc   +1 more source

Rank‐based estimation of propensity score weights via subclassification

Canadian Journal of Statistics, EarlyView.
Abstract Propensity score (PS) weighting estimators are widely used for causal effect estimation and enjoy desirable theoretical properties, such as consistency and potential efficiency under correct model specification. However, their performance can degrade in practice due to sensitivity to PS model misspecification.
Linbo Wang, Yuexia Zhang, Thomas S. Richardson, Xiaohua Zhou   +3 more
wiley   +1 more source

Subuniformity of harmonic mean p$$ p $$‐values

Canadian Journal of Statistics, EarlyView.
Abstract We obtain several inequalities on the generalized means of dependent p$$ p $$‐values. In particular, the weighted harmonic mean of p$$ p $$‐values is strictly subuniform under several dependence assumptions of p$$ p $$‐values, including independence, negative upper orthant dependence, the class of extremal mixture copulas, and some Clayton ...
Yuyu Chen, Ruodu Wang, Yuming Wang, Wenhao Zhu   +3 more
wiley   +1 more source

On subset least squares estimation and prediction in vector autoregressive models with exogenous variables

Canadian Journal of Statistics, EarlyView.
Abstract We establish the consistency and the asymptotic distribution of the least squares estimators of the coefficients of a subset vector autoregressive process with exogenous variables (VARX). Using a martingale central limit theorem, we derive the asymptotic normal distribution of the estimators. Diagnostic checking is discussed using kernel‐based
Pierre Duchesne, Thanh‐Liem Huynh, Mylène Oscar   +2 more
wiley   +1 more source

T‐calibration in semi‐parametric models

Canadian Journal of Statistics, EarlyView.
Abstract This article relates the calibration of models to the consistent loss functions for the target functional of the model. Correctly specified models are calibrated. Conversely, we demonstrate that if there is a parameter value that is optimal under all consistent loss functions, then a model is calibrated.
Anja Mühlemann, Johanna Ziegel
wiley   +1 more source

High quantile function estimation

, 2003
L'objectif de ce travail est l'estimation d'une fonction quantile extrême. Nous considérons n couples de variables aléatoires indépendants et de même loi qu'un couple (X,Y) à support borné du plan. Notre but est d'estimer le quantile extrême (i.e. d'ordre inférieur à 1/n) de la fonction de répartition conditionnelle de Y sachant que X=x. La fonction de
openaire   +1 more source

Asymptotic independence in more than two dimensions and its implications on risk management

Canadian Journal of Statistics, EarlyView.
Abstract In extreme value theory, the presence of asymptotic independence signifies that joint extreme events across multiple variables are unlikely. Although well understood in a bivariate context, the concept remains relatively unexplored when addressing the nuances of simultaneous occurrence of extremes in higher dimensions.
Bikramjit Das, Vicky Fasen‐Hartmann
wiley   +1 more source