首页|University of Warwick Reports Findings in Machine Learning (Assessment of machin e learning models trained by molecular dynamics simulations results for inferrin g ethanol adsorption on an aluminium surface)
University of Warwick Reports Findings in Machine Learning (Assessment of machin e learning models trained by molecular dynamics simulations results for inferrin g ethanol adsorption on an aluminium surface)
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 Daily News – New research on Machine Learning is the subject o f a report. According to news originating from Coventry, United Kingdom, by News Rx correspondents, research stated, “Molecular dynamics (MD) simulations can red uce our need for experimental tests and provide detailed insight into the chemic al reactions and binding kinetics. There are two challenges while dealing with M D simulations: one is the time and length scale limitations, and the latter is e fficiently processing the massive amount of data resulting from the MD simulatio ns and generating the proper reaction rates.” Our news journalists obtained a quote from the research from the University of W arwick, “In this work, we evaluated the use of regression machine learning (ML) methods to solve these two challenges by developing a framework for ethanol adso rption on an Aluminium (Al) slab. This framework comprises three main stages: fi rst, an all-atom molecular dynamics model; second, ML regression models; and thi rd, validation and testing. In stage one, the adsorption of ethanol molecules on the Al surface for various temperatures, velocities and concentrations is simul ated using the large-scale atomic/molecular massively parallel simulator (LAMMPS ) and ReaxFF. The outcome of stage one is utilised for training, testing, and va lidating the predictive models in stages two and three. We developed and evaluat ed 28 different ML models for predicting the number of adsorbed molecules over t ime, including linear regression, support vector machine (SVM), decision trees, ensemble, Gaussian process regression (GPR), neural network (NN) and Bayesian hy per-parameter optimisation models. Based on the results, the Bayesian-based GPR showed the highest accuracy and the lowest training time. The developed model ca n predict the number of adsorbed molecules for new cases within seconds, while M D simulations take a few weeks. This adsorption rate can then be used in macrosc ale simulations to tackle the time and length scale limitations.”
NETL
NSTL
