Enhancing Classical Simulations with Electronic Corrections and Artifical Neural Networks.

Description

Classical simulations of materials and nanoparticles have the advantage of speed and scalability, but lack the ability to describe the electronic proeprties.  Quantum mechanical and ab initio simulations have the advantage of providing accurate estimations of the electronic structure and charge transfer, but are typically limited to small and simple systems.  In this project identical sets of classical and electronic structure simualtions will be used to train an artificial neural network to predict a correction term, and the Fermi energy of arbitary gold nanoparticles.  This 6 credit point project involves some scientific programming to characterise the data and extract suitable features, data science and simple introductory machine learning.

Requirements

coding in C++ and python, good mathematics and statistics, interest in machine learning

Keywords

nanoparticles, materials science, simulation, scientific programming, computational science, machine learning, neural networks

Updated:  1 June 2019/Responsible Officer:  Dean, CECS/Page Contact:  CECS Marketing