Results for “KSe” | Engineering Papers

DOE OSTI2020

Materials Data on KSe by Materials Project

KSe crystallizes in the hexagonal P-62m space group. The structure is three-dimensional. there are two inequivalent K1+ sites. In the first K1+ site, K1+ is bonded in a 6-coordinate geometry to six Se1- atoms. There are two shorter (3.39 Å) and four longer (3.46 Å) K–Se bond lengths. In the second K1+ site, K1+ is bonded in a 6-coordinate geometry to six Se1- atoms. There are four shorter (3.32 Å) and two longer (3.39 Å) K–Se bond lengths. There are two inequivalent Se1- sites. In the first Se1- site, Se1- is bonded in a 7-coordinate geometry to six K1+ and one Se1- atom. The Se–Se bond length is 2.45 Å. In the second Se1- site, Se1- is bonded in a 7-coordinate geometry to six K1+ and one Se1- atom. The Se–Se bond length is 2.44 Å.

36 MATERIALS SCIENCE↗

DOE OSTI2023

New machine learning techniques for simulation-based inference: InferoStatic nets, kernel score estimation, and kernel likelihood ratio estimation

We propose an intuitive, machine-learning approach to multiparameter inference, dubbed the InferoStatic Networks (ISN) method, to model the score and likelihood ratio estimators in cases when the probability density can be sampled but not computed directly. The ISN uses a backend neural network that models a scalar function called the inferostatic potential \varphi φ . In addition, we introduce new strategies, respectively called Kernel Score Estimation (KSE) and Kernel Likelihood Ratio Estimation (KLRE), to learn the score and the likelihood ratio functions from simulated data. We illustrate the new techniques with some toy examples and compare to existing approaches in the literature. We mention en passant some new loss functions that optimally incorporate latent information from simulations into the training procedure.

Kong, Kyoungchul↗

DOE OSTI2023

Codebase release 0.1 for infstat

We propose an intuitive, machine-learning approach to multiparameter inference, dubbed the InferoStatic Networks (ISN) method, to model the score and likelihood ratio estimators in cases when the probability density can be sampled but not computed directly. The ISN uses a backend neural network that models a scalar function called the inferostatic potential \varphi φ . In addition, we introduce new strategies, respectively called Kernel Score Estimation (KSE) and Kernel Likelihood Ratio Estimation (KLRE), to learn the score and the likelihood ratio functions from simulated data. We illustrate the new techniques with some toy examples and compare to existing approaches in the literature. We mention en passant some new loss functions that optimally incorporate latent information from simulations into the training procedure.

Kong, Kyoungchul↗

NASA NTRS1998

Every Breath You Take

As a result of SBIR (Small Business Innovative Research) work for Johnson Space Center, KSE, Inc., developed a photocatalytic Adsorption-Integrated-Reaction (AIR) air purification process originally used for air revitalization in life support systems during extended space operations. A unit has been purchased by a commercial customer for control of emissions of airborne chlorinated hydrocarbons at a Superfund site.

Source record↗

DOE OSTI2022

New Machine Learning Techniques for Simulation-Based Inference: InferoStatic Nets, Kernel Score Estimation, and Kernel Likelihood Ratio Estimation

We propose an intuitive, machine-learning approach to multiparameter inference, dubbed the InferoStatic Networks (ISN) method, to model the score and likelihood ratio estimators in cases when the probability density can be sampled but not computed directly. The ISN uses a backend neural network that models a scalar function called the inferostatic potential $\varphi$. In addition, we introduce new strategies, respectively called Kernel Score Estimation (KSE) and Kernel Likelihood Ratio Estimation (KLRE), to learn the score and the likelihood ratio functions from simulated data. We illustrate the new techniques with some toy examples and compare to existing approaches in the literature. We mention en passant some new loss functions that optimally incorporate latent information from simulations into the training procedure.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗