The objective of this project is to use deep neural networks to rapidly approximate difficult and computationally demanding problems in material simulation, enabling the "bottom up" approach to nanoscale materials and device design that has long been promised, but so far remains unrealized. We will use recent advances in artificial intelligence and deep learning to solve applied problems in physics, chemistry, and engineering.
The project will investigate application of DNN to solving the electronic structure problem (within the density functional theory and wavefunction based methods). The objective is to show that DNN can outperform "traditional" electronic structure methods by a factor of 10,000. We will generate predictive results 100 times faster than is currently possible and work on problems 100 times larger than can currently be modelled.
This project represents a fundamentally new approach to numerical computation and simulation.
Education: PhD in a computational-based discipline such as electrical or software engineering, signal processing, computational physics, computer science, or equivalent. The candidates will have demonstrated proficiency in the field of applied deep learning, working with large datasets using deep learning frameworks such as TensorFlow, Theano, or Caffe.
Salary = $64,000 CAD / year for two years
Location = Ottawa, Ontario, Canada
Coffee = free
Candidates should have obtained a PhD (or equivalent) within the past three years (PhD received on or after July 1, 2014) or expect to complete their PhD by September, 2017. To be eligible, a candidate will be required to submit their CV, three publications, a statement of interest, and send the names of three references with contact information.
Application deadline = 22 April 2017