Filters








61,959 Hits in 8.3 sec

MODELING NUCLEAR DATA UNCERTAINTIES USING DEEP NEURAL NETWORKS

Majdi I. Radaideh, Dean Price, Tomasz Kozlowski, M. Margulis, P. Blaise
2021 EPJ Web of Conferences  
A new concept using deep learning in neural networks is investigated to characterize the underlying uncertainty of nuclear data.  ...  Although of the very limited sample size (1000 samples) available in this study, the trained models demonstrate promising performance, where a prediction error of about 166 pcm is found for keff in the  ...  Next, Sampler receives the "pre-determined" cross-section libraries and uses them for forward uncertainty propagation to determine the effect of nuclear data uncertainty on the output.  ... 
doi:10.1051/epjconf/202124715016 fatcat:pg7dtgxbmfdyjijqizkkirjyom

Deep Horizon; a machine learning network that recovers accreting black hole parameters [article]

Jeffrey van der Gucht, Jordy Davelaar, Luc Hendriks, Oliver Porth, Hector Olivares, Yosuke Mizuno, Christian M. Fromm, Heino Falcke
2019 arXiv   pre-print
We train two convolutional deep neural networks on a large image library of simulated mock data.  ...  We investigate the effects of a limited telescope resolution and of observations at different frequencies.  ...  This work was funded by the ERC Synergy Grant "BlackHoleCam-Imaging the Event Horizon of Black Holes" (Grant 610058, Goddi et al. (2017) ). The  ... 
arXiv:1910.13236v1 fatcat:hry74ga5wfgqtelwrenzesissm

Deep Horizon: A machine learning network that recovers accreting black hole parameters

Jeffrey van der Gucht, Jordy Davelaar, Luc Hendriks, Oliver Porth, Hector Olivares, Yosuke Mizuno, Christian M. Fromm, Heino Falcke
2020 Astronomy and Astrophysics  
We trained two convolutional deep neural networks on a large image library of simulated mock data.  ...  We investigate the effects of a limited telescope resolution and observations at higher frequencies. Methods.  ...  This work was funded by the ERC Synergy Grant A94, page 11 of 12 A&A 636, A94 (2020) "BlackHoleCam-Imaging the Event Horizon of Black Holes" (Grant 610058, Goddi et al. 2017) .  ... 
doi:10.1051/0004-6361/201937014 fatcat:4vyvbfxdlzbrdgppicmvtumgwm

Probabilistic Deep Learning with Probabilistic Neural Networks and Deep Probabilistic Models [article]

Daniel T. Chang
2021 arXiv   pre-print
Probabilistic deep learning is deep learning that accounts for uncertainty, both model uncertainty and data uncertainty. It is based on the use of probabilistic models and deep neural networks.  ...  TensorFlow Probability is a library for probabilistic modeling and inference which can be used for both approaches of probabilistic deep learning. We include its code examples for illustration.  ...  Then, one has to choose a probability model: a prior distribution over the possible model parameters p(θ) and a prior confidence in the predictive power of the model p(y | x, θ), i.e., BNN θ (x).  ... 
arXiv:2106.00120v3 fatcat:gbeonxch4vav7jaqu3nvti7thi

Estimating atmospheric parameters from LAMOST low-resolution spectra with low SNR

Xiangru Li, Si Zeng, Zhu Wang, Bing Du, Xiao Kong, Caixiu Liao
2022 Monthly notices of the Royal Astronomical Society  
The effectiveness of the LASSO-MLPNet was evaluated on some LAMOST stellar spectra of the common star between APOGEE (The Apache Point Observatory Galactic Evolution Experiment) and LAMOST. it is shown  ...  Finally, LASSO-MLPNet used a Multilayer Perceptron network (MLPNet) to estimate atmospheric parameters Teff, log g and [Fe/H].  ...  LAMOST, a multi-target optical fiber spectroscopic telescope in the large sky area, is a major national engineering project built by the Chinese Academy of Sciences.  ... 
doi:10.1093/mnras/stac1625 fatcat:ar2f2ygxirfkpb7iulmh3mrls4

Inferring Structural Parameters of Low-Surface-Brightness-Galaxies with Uncertainty Quantification using Bayesian Neural Networks [article]

Dimitrios Tanoglidis, Aleksandra Ćiprijanović, Alex Drlica-Wagner
2022 arXiv   pre-print
In this work, we demonstrate that a Bayesian Neural Network (BNN) can be used for the inference, with uncertainty quantification, of such morphological parameters from simulated low-surface-brightness  ...  Measuring the structural parameters (size, total brightness, light concentration, etc.) of galaxies is a significant first step towards a quantitative description of different galaxy populations.  ...  We acknowledge the Deep Skies Lab as a community of multi-domain experts and collaborators who've facilitated an environment of open discussion, idea-generation, and collaboration.  ... 
arXiv:2207.03471v1 fatcat:6mvrd5suxvbijmr5c2fy4dxwdu

Epistemic Neural Networks [article]

Ian Osband, Zheng Wen, Seyed Mohammad Asghari, Vikranth Dwaracherla, Morteza Ibrahimi, Xiuyuan Lu, Benjamin Van Roy
2022 arXiv   pre-print
While prior approaches to uncertainty modeling such as Bayesian neural networks can be expressed as ENNs, this new interface facilitates comparison of joint predictions and the design of novel architectures  ...  We introduce the epistemic neural network (ENN) as an interface for models that represent uncertainty as required to generate useful joint predictions.  ...  and code with the Uncertainty Baselines project (Nado et al., 2021) .  ... 
arXiv:2107.08924v5 fatcat:gaiua6m5vbckvbpmsv2vysxd2u

TyXe: Pyro-based Bayesian neural nets for Pytorch [article]

Hippolyt Ritter, Theofanis Karaletsos
2021 arXiv   pre-print
We introduce TyXe, a Bayesian neural network library built on top of Pytorch and Pyro.  ...  We showcase the ease of use of TyXe to explore Bayesian versions of popular models from various libraries: toy regression with a pure Pytorch neural network; large-scale image classification with torchvision  ...  We then leverage Pyro to formulate a probabilistic model over the neural network parameters, in which we perform approximate inference.  ... 
arXiv:2110.00276v1 fatcat:x6f37xqiefcwpjh6qhvscx6shu

SDSS-IV MaStar: Data-driven Parameter Derivation for the MaStar Stellar Library [article]

Julie Imig, Jon A. Holtzman, Renbin Yan, Daniel Lazarz, Yanping Chen, Lewis Hill, Daniel Thomas, Claudia Maraston, Moire M. K. Prescott, Guy S. Stringfellow, Dmitry Bizyaev, Rachael L. Beaton (+1 others)
2021 arXiv   pre-print
These parameters are derived with a flexible data-driven algorithm that uses a neural network model.  ...  We present our derived parameters along with an analysis of the uncertainties and comparisons to other analyses from the literature.  ...  We train a neural network on this set to reproduce a spectrum as a function of its parameters, then use the network to determine the parameters of the remaining spectra in the library.  ... 
arXiv:2112.01669v1 fatcat:jyjvvdtulbcx7ms44dibwcmtj4

SDSS-IV MaStar: Data-driven Parameter Derivation for the MaStar Stellar Library

Julie Imig, Jon A. Holtzman, Renbin Yan, Daniel Lazarz, Yanping Chen, Lewis Hill, Daniel Thomas, Claudia Maraston, Moire M. K. Prescott, Guy S. Stringfellow, Dmitry Bizyaev, Rachael L. Beaton (+1 others)
2022 Astronomical Journal  
These parameters are derived with a flexible data-driven algorithm that uses a neural network model.  ...  We present our derived parameters along with an analysis of the uncertainties and comparisons to other analyses from the literature.  ...  ORCID iDs Julie Imig https:/ /orcid.org/0000-0003-2025-3585 Jon A.  ... 
doi:10.3847/1538-3881/ac3ca7 fatcat:v5hjzjgvwrbazaqz4qe3uqdu3u

Extrapolative Bayesian Optimization with Gaussian Process and Neural Network Ensemble Surrogate Models

Yee-Fun Lim, Chee Koon Ng, U.S. Vaitesswar, Kedar Hippalgaonkar
2021 Advanced Intelligent Systems  
It is believed that these findings shed new light on the understanding of surrogate models for BO, and can help accelerate the inverse design of new materials with better structural and functional performance  ...  the second is a simulated high-dimensional dataset of thermoelectric properties of inorganic materials (22 inputs and 1 target).  ...  This work was supported by the Agency of Science, Technology and Research (A*STAR), Singapore, via two programmatic research grants (grant nos. A1898b0043 and A19E9a0103 ).  ... 
doi:10.1002/aisy.202100101 fatcat:tnmlfhgsyven5lgjywk7oexpdq

Addressing biological uncertainties in engineering gene circuits

Carolyn Zhang, Ryan Tsoi, Lingchong You
2016 Integrative Biology  
We discuss biological uncertainties that complicate predictable engineering of gene circuits and potential strategies to address these uncertainties.  ...  Quantifying effects of parts on the chassis 2.1 Computational models. Modeling can evaluate the impact of uncertainty on circuit function and the robustness of a circuit against these uncertainties.  ...  Specific network motifs can resist effects of noise on circuit function (Fig. 2D ).  ... 
doi:10.1039/c5ib00275c pmid:26674800 pmcid:PMC4837052 fatcat:y7ae6c2jufddxn4l4yfqagluly

BNNpriors: A library for Bayesian neural network inference with different prior distributions

Vincent Fortuin, Adrià Garriga-Alonso, Mark van der Wilk, Laurence Aitchison
2021 Software Impacts  
Our new library, BNNpriors, enables state-of-the-art Markov Chain Monte Carlo inference on Bayesian neural networks with a wide range of predefined priors, including heavy-tailed ones, hierarchical ones  ...  It has facilitated foundational discoveries on the nature of the cold posterior effect in Bayesian neural networks and will hopefully catalyze future research as well as practical applications in this  ...  Acknowledgments VF was supported by a PhD fellowship from the Swiss Data Science Center and by the grant #2017-110 of the Strategic Focus Area ''Personalized Health and Related Technologies (PHRT)'' of  ... 
doi:10.1016/j.simpa.2021.100079 fatcat:4da5sth5tbfkpch2chxsd4re2a

Real-time parameter inference in reduced-order flame models with heteroscedastic Bayesian neural network ensembles [article]

Ushnish Sengupta, Maximilian L. Croci, Matthew P. Juniper
2020 arXiv   pre-print
We train our networks on a library of 2.1 million simulated flame videos.  ...  Results on the test dataset of simulated flames show that the network recovers flame model parameters, with the correlation coefficient between predicted and true parameters ranging from 0.97 to 0.99,  ...  Acknowledgments and Disclosure of Funding This project has received funding from the European Union's Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement number  ... 
arXiv:2011.02838v1 fatcat:dnes2apumjfs7okrvzaht3ha24

MODELLING OF EQUIPMENT FAILURE RATE ACCOUNTING FOR THE UNCERTAINTY

H.X. Tian, W.F Wu, P. Wang, H.Z. Li
2015 International Journal on Smart Sensing and Intelligent Systems  
A fuzzy model for failure rate with the consideration of the effects of uncertain factors in distribution reliability evaluation is presented.  ...  The possibility and credibility distribution analyzed on the basis of sample datum are used for quantifying effects of the uncertainty done to failure rate.  ...  Conclusions A novel model of the equipment failure rate for incorporating and weigh the effects of uncertainty factors in distribution network is proposed.  ... 
doi:10.21307/ijssis-2017-816 fatcat:bm4em2lhirgovl4txy2jl3e6vq
« Previous Showing results 1 — 15 out of 61,959 results