a•kin’s Chief Research Officer, Alexander Hadjiivanov, recently won the Outstanding Student Paper Award at the 2019 IEEE Congress on Evolutionary Computation (CEC 2019) in Wellington, New Zealand, in stiff competition against hundreds of other PhD candidates from around the world. 

Alex’s paper ‘Epigenetic evolution of deep convolutional models’ (to appear on IEEEXplore shortly) is about a hybrid optimisation strategy where evolution is used for neural architecture search and backpropagation is used for updating the parameters of a convolutional neural network. 

The study also introduces a new mutation operator for convolutional kernels which can be used to evolve irregularly shaped kernels within the same convolutional layer. Alex submitted this paper to the conference in tandem with submitting his PhD entitled ‘Biologically inspired evolution and adaptation in artificial neural networks’.

Image credit: Mandy Simpson [CC BY-SA 3.0 (http://creativecommons.org/licenses/by-sa/3.0/)]