Abstract
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 reporting out of Los Angeles, California, by NewsR x editors, research stated, “Unraveling the reaction pathway of photoinduced rea ctions poses a great challenge owing to its complexity. Recently, graph theory-b ased machine learning combined with nonadiabatic molecular dynamics (NAMD) has b een applied to obtain the global reaction coordinate of the photoisomerization o f azobenzene.” Our news journalists obtained a quote from the research from the University of S outhern California, “However, NAMD simulations are computationally expensive as they require calculating the nonadiabatic coupling vectors at each time step. He re, we showed that ab initio molecular dynamics (AIMD) can be used as an alterna tive to NAMD by choosing an appropriate initial condition for the simulation. We applied our methodology to determine a plausible global reaction coordinate of retinal photoisomerization, which is essential for human vision. On rank-orderin g the internal coordinates, based on the mutual information (MI) between the int ernal coordinates and the HOMO energy, NAMD and AIMD give a similar trend.”
NETL
NSTL