首页|Northeastern University Reports Findings in Machine Learning (Uncertainty Quanti fication of Linear Scaling, Machine Learning, and Density Functional Theory Deri ved Thermodynamics for the Catalytic Partial Oxidation of Methane on Rhodium)
Northeastern University Reports Findings in Machine Learning (Uncertainty Quanti fication of Linear Scaling, Machine Learning, and Density Functional Theory Deri ved Thermodynamics for the Catalytic Partial Oxidation of Methane on Rhodium)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews-New research on Machine Learning is the subject o f a report. According to news originating fromBoston, Massachusetts, by NewsRx correspondents, research stated, "Accurate and complete microkineticmodels (MKM s) are powerful for anticipating the behavior of complex chemical systems at dif ferentoperating conditions. In heterogeneous catalysis, they can be further use d for the rapid development andscreening of new catalysts."Our news journalists obtained a quote from the research from Northeastern Univer sity, "Densityfunctional theory (DFT) is often used to calculate the parameters used in MKMs with relatively highfidelity. However, given the high cost of DFT calculations for adsorbates in heterogeneous catalysis, linearscaling relation s (LSRs) and machine learning (ML) models were developed to give rapid estimates of theparameters in MKM. Regardless of the method, few studies have attempted to quantify the uncertainty incatalytic MKMs, as the uncertainties are often or ders of magnitude larger than those for gas phase models.This study explores un certainty quantification and Bayesian Parameter Estimation for thermodynamic par ameters calculated by DFT, LSRs, and GemNet-OC, a ML model developed under the O pen CatalystProject. A model for catalytic partial oxidation of methane (CPOX) on Rhodium was chosen as a casestudy, in which the model's thermodynamic parame ters and their associated uncertainties were determinedusing DFT, LSR, and GemN et-OC. Markov Chain Monte Carlo coupled with Ensemble Slice Sampling wasused to sample the highest probability density (HPD) region of the posterior and determ ine the maximumof the a posteriori (MAP) for each thermodynamic parameter inclu ded. The optimized microkineticmodels for each of the three estimation methods had quite similar mechanisms and agreed well with theexperimental data for gas phase mole fractions. Exploration of the HPD region of the posterior further revealed that adsorbed hydroxide and oxygen likely bind on facets other than Rhodiu m 111."
BostonMassachusettsUnited StatesNo rth and Central AmericaAlkanesCyborgsEmerging TechnologiesMachine Learni ngMethanePhysicsRhodiumThermodynamicsTransition Elements
NETL
NSTL
