Results 121 to 130 of about 113,685 (321)

Ensemble Kalman filter in latent space using a variational autoencoder pair

Quarterly Journal of the Royal Meteorological Society, EarlyView.
The use of the ensemble Kalman filter (EnKF) in strongly nonlinear or constrained atmospheric, oceanographic, or sea‐ice models can be challenging. Applying the EnKF in the latent space of a variational autoencoder (VAE) ensures that the ensemble members satisfy the balances and constraints present in the model.
Ivo Pasmans, Yumeng Chen, Tobias Sebastian Finn, Marc Bocquet, Alberto Carrassi   +4 more
wiley   +1 more source

Data assimilation with extremum Monte Carlo methods

Quarterly Journal of the Royal Meteorological Society, EarlyView.
This study presents the extremum Monte Carlo filter as a data assimilation method and, in particular, a variant of the variational approach (three‐ and four‐dimensional variational), where the state estimates are obtained by solving an optimization problem numerically over a space of prediction functions, instead of the state space itself.
Karim Moussa, Siem Jan Koopman
wiley   +1 more source

Improved Quantile Convolutional and Recurrent Neural Networks for Electric Vehicle Battery Temperature Prediction

Big Data Mining and Analytics
The battery thermal management of electric vehicles can be improved using neural networks predicting quantile sequences of the battery temperature. This work extends a method for the development of Quantile Convolutional and Quantile Recurrent Neural ...
Andreas M. Billert, Runyao Yu, Stefan Erschen, Michael Frey, Frank Gauterin   +4 more
doaj   +1 more source

An objective Bayesian method for including parameter uncertainty in ensemble model output statistics

Quarterly Journal of the Royal Meteorological Society, EarlyView.
Conventional model output statistics and ensemble model output statistics methods for calibrating ensemble forecasts lead to severe underestimation of the probabilities of ensemble extremes (in blue). This is because they ignore statistical parameter uncertainty.
Stephen Jewson, Leonardo Olivetti, Gabriele Messori, Paul J. Northrop, Trevor Sweeting   +4 more
wiley   +1 more source

Identify Relative importance of covariates in Bayesian lasso quantile regression via new algorithm in statistical program R [PDF]

Revista Română de Statistică, 2017
In this paper, we propose a new algorithm to determine the relative importance of covariates by Bayesian Lasso quantile regression for variable selection assigning new formula of Laplace distributions for the regression parameters.
Fadel Hamid Hadi Alhusseini, Taha al Shaybawee, Fedaa Abd Almajid Sabbar Alaraje   +2 more
doaj  

A composite‐loss graph neural network for the multivariate post‐processing of ensemble weather forecasts

Quarterly Journal of the Royal Meteorological Society, EarlyView.
The dual graph neural network (dualGNN), trained with a composite loss combining the energy score (ES) and variogram score (VS), consistently outperformed models optimized solely for ES or the continuous ranked probability score in the multivariate setting, as well as empirical copula approaches.
Mária Lakatos
wiley   +1 more source

COBASE: A new copula‐based shuffling method for ensemble weather forecast postprocessing

Quarterly Journal of the Royal Meteorological Society, EarlyView.
We propose COBASE, a novel copula‐based postprocessing methododology that combines the strengths of multivariate parametric correction with non‐parametric rank‐based approaches. We consider two case studies for multi‐site temperature in Austria and multi‐site temperature and dew‐point temperature in the Netherlands.
Maurits Flos, Bastien François, Irene Schicker, Kirien Whan, Elisa Perrone   +4 more
wiley   +1 more source

Bayesian Endogenous Tobit Quantile Regression

Bayesian Analysis, 2017
This study proposes $p$-th Tobit quantile regression models with endogenous variables. In the first stage regression of the endogenous variable on the exogenous variables, the assumption that the $ $-th quantile of the error term is zero is introduced.
openaire   +3 more sources

Bayesian quantile regression with approximate likelihood

Bernoulli, 2015
Quantile regression is often used when a comprehensive relationship between a response variable and one or more explanatory variables is desired. The traditional frequentists' approach to quantile regression has been well developed around asymptotic theories and efficient algorithms.
Feng, Yang, Chen, Yuguo, He, Xuming
openaire   +3 more sources

Physics‐Informed Neural Networks for Battery Degradation Prediction Under Random Walk Operations

Quality and Reliability Engineering International, EarlyView.
ABSTRACT This study addresses the challenge of predicting the state of health (SoH) and capacity degradation in Battery Energy Storage Systems (BESS) under highly variable conditions induced by frequent control adjustments. In environments where random walk behavior prevails due to stochastic control commands, conventional estimation methods often ...
Alaa Selim, Huadong Mo, Hemanshu Pota, Daoyi Dong   +3 more
wiley   +1 more source