111 research outputs found
Randa-lab/Bayesian_Neural_Network_Probabilistic_Ionosphere_VTEC: Bayesian_Neural_Network_Probabilistic_Ionosphere
No full textBayesian neural network models for probabilistic VTEC forecasting with 95% confidence, from the paper "Uncertainty Quantification for Machine Learning-based Ionosphere and Space Weather Forecasting" by Natras Randa et al., submitted to the Space Weather Jornal, AGU.
Two Bayesian neural network models were developed for VTEC forecasting with 95% confidence intervals. In both models, the deterministic network parameters (weights) are replaced by probability distributions of these weights. The first model only describes the uncertainty in the weights and estimates the model uncertainty, while the second model also estimate the data uncertainty by providing probabilistic output estimate via minimizing the negative log-likelihood (NLL) loss
Randa-lab/Bayesian_Neural_Network_Probabilistic_Ionosphere_VTEC: Bayesian_Neural_Network_Probabilistic_Ionosphere
No full textBayesian neural network models for probabilistic VTEC forecasting with 95% confidence, from the paper "Uncertainty Quantification for Machine Learning-based Ionosphere and Space Weather Forecasting" by Natras Randa et al., submitted to the Space Weather Jornal, AGU.
Two Bayesian neural network models were developed for VTEC forecasting with 95% confidence intervals. In both models, the deterministic network parameters (weights) are replaced by probability distributions of these weights. The first model only describes the uncertainty in the weights and estimates the model uncertainty, while the second model also estimate the data uncertainty by providing probabilistic output estimate via minimizing the negative log-likelihood (NLL) loss
Randa-lab/Quantile_Gradient_Boosting_for_Probabilistic_VTEC: Quantile_Gradient_Boosting_Probabilistic_Ionosphere_Evaluation
No full textThis release demonstrates how to load and evaluate the probabilistic Quantile Gradient Boosting (QGB) Vertical Total Electron Content (VTEC) models, which provide 95% confidence intervals. QGB VTEC models forecast VTEC 1-day ahead for grid points at 10° of longitude, and 10°, 40°, and 70° of latitude. They were developed within the study "Uncertainty Quantification for Machine Learning-based Ionosphere and Space Weather Forecasting" by Natras R., Soja B. and Schmidt M., submitted to the Space Weather Jornal, AGU.
Quantiles were estimated by multiplying the quantile values β by the positive and negative residuals in the loss function to obtain the quantile loss (QL) (see Equation 7 in the paper). Quantile values of β ={0.025, 0.975} are chosen for estimating the lower and upper confidence bounds, respectively, to obtain a confidence interval of 95%. The mean quantile β={0.50} provides the median VTEC
The Scope of Themes of the Story “Randa” by Zakaria Tamer and its Series of Images
No full textThis article is devoted to the themes and figurative system of the story of the Syrian writer Zakaria Tamer. The author combined several relevant topics in one story and, with the help of the main character of the story, the girl Randa, gathered in one image all the children of the Syrian people, in particular, the entire Arab people
Latitudinal and Seasonal Dependence of Geomagnetic Storm Effects on the Ionospheric TEC and Positioning Accuracy in Europe in 2015
No full text
Strong solar flare detection and its impact on ionospheric layers and on coordinates accuracy in the Western Balkans in October 2014
No full textUncertainty Quantification for Machine Learning‐Based Ionosphere and Space Weather Forecasting: Ensemble, Bayesian Neural Network, and Quantile Gradient Boosting
No full textAbstractMachine learning (ML) has been increasingly applied to space weather and ionosphere problems in recent years, with the goal of improving modeling and forecasting capabilities through a data‐driven modeling approach of nonlinear relationships. However, little work has been done to quantify the uncertainty of the results, lacking an indication of how confident and reliable the results of an ML system are. In this paper, we implement and analyze several uncertainty quantification approaches for an ML‐based model to forecast Vertical Total Electron Content (VTEC) 1‐day ahead and corresponding uncertainties with 95% confidence intervals (CI): (a) Super‐Ensemble of ML‐based VTEC models (SE), (b) Gradient Tree Boosting with quantile loss function (Quantile Gradient Boosting, QGB), (c) Bayesian neural network (BNN), and (d) BNN including data uncertainty (BNN + D). Techniques that consider only model parameter uncertainties (a and c) predict narrow CI and over‐optimistic results, whereas accounting for both model parameter and data uncertainties with the BNN + D approach leads to a wider CI and the most realistic uncertainties quantification of VTEC forecast. However, the BNN + D approach suffers from a high computational burden, while the QGB approach is the most computationally efficient solution with slightly less realistic uncertainties. The QGB CI are determined to a large extent from space weather indices, as revealed by the feature analysis. They exhibit variations related to daytime/nightime, solar irradiance, geomagnetic activity, and post‐sunset low‐latitude ionosphere enhancement.Plain Language Summary: Space weather describes the varying conditions in the space environment between the Sun and Earth that can affect satellites and technologies on Earth, such as navigation systems, power grids, radio, and satellite communications. The manifestation of space weather in the ionosphere can be characterized using the Vertical Total Electron Content (VTEC) derived from Global Navigation Satellite Systems observations. In this study, the machine learning (ML) approach is applied to approximate the nonlinear relationships of Sun‐Earth processes using data on solar activity, solar wind, magnetic field, and VTEC. However, the measurements and the modeling approaches are subject to errors, increasing the uncertainty of the results when forecasting future instances. For reliable forecasting, it is necessary to quantify the uncertainties. Quantifying the uncertainty is also helpful for understanding the ML‐based model and the problem of VTEC and space weather forecasting. Therefore, in this study, ML‐based models are developed to forecast VTEC within the ionosphere, including the manifestation of space weather, while the degree of reliability is quantified with a target value of 95% confidence.Key Points:
Machine learning‐based Vertical Total Electron Content models with 95% confidence intervals (CI) are developed for the first time using four approaches to quantify uncertainties
Bayesian Neural Network quantifying model and data uncertainties contains ground truth within CIs, but is computationally intensive
Quantile Gradient Boosting is fastest with comparable performance in terms of uncertainty; CIs largely determined from space weather indices
Deutscher Akademischer Austauschdienst
http://dx.doi.org/10.13039/501100001655https://www.tensorflow.org/https://doi.org/10.21105/joss.03021http://www.aiub.unibe.ch/download/CODEhttps://kauai.ccmc.gsfc.nasa.gov/instantrun/irihttps://doi.org/10.5281/zenodo.7741342https://doi.org/10.5281/zenodo.7858906https://doi.org/10.5281/zenodo.785866
Strong Solar Flares Detection, Its Impact to D and F Ionospheric Regions and to Coordinates Accuracy in Western Balkan in October 2014
No full text- …
