Andrew M. Webb
I am a research associate in the computer science department of the University of Manchester, currently working on the LAMBDA project under Gavin Brown, studying methods of training deep neural network ensembles for regression and classification that reward model diversity, with a sub-goal of creating efficient deep neural networks. I've also worked recently on the PAMELA project, working on the efficient integration of semantic segmentation into state-of-the-art SLAM (simultaneous localisation and mapping) systems.
My PhD research was on selection for evolvability in evolutionary algorithms, where evolvability is defined as the capacity for adaptive evolution. I've also done some work in artificial life, studying self-reproducing automata.
I maintain QCircuits (), a Python package for simulating the execution of algorithms on small-scale quantum computers.
- Andrew M. Webb, Charles Reynolds, et al. (2019). Joint Training of Neural Network Ensembles, preprint, arXiv:1902.04422
- Sajad Saeedi, Bruno Bodin, et al. (2018). Navigating the Landscape for Real-Time Localization and Mapping for Robotics and Virtual and Augmented Reality, Proceedings of the IEEE Vol. 106
- Andrew M. Webb (2016). On Selection for Evolvability. PhD thesis. School of Computer Science, University of Manchester, UK. (Supervisors: Prof. Joshua Knowles and Dr Julia Handl. Examiners: Dr Simon Powers and Dr Jonathan Shapiro.)
- Andrew M. Webb, Julia Handl, and Joshua Knowles (2015). How Much Should You Select for Evolvability?, Proceedings of the European Conference on Artificial Life (ECAL) 2015, MIT Press
- Andrew M. Webb and Joshua Knowles (2014). Studying the Evolvability of Self-Encoding Genotype-Phenotype Maps, Proceedings of the International Conference on the Synthesis and Simulation of Living Systems (ALIFE) 2014, MIT Press
- Andrew M. Webb, Sergio Davies, and David Lester (2011). Spiking Neural PID Controllers, International Conference on Neural Information Processing