Alan J. Lockett
Revision as of 20:31, 13 July 2020 by GerdIsenberg (talk | contribs) (Created page with "'''Home * People * Alan J. Lockett''' FILE:alan-lockett.jpg|border|right|thumb|link=http://nn.cs.utexas.edu/?alanlockett| Alan J. Lockett <ref>[http://nn....")
Home * People * Alan J. Lockett
Alan J. Lockett,
an American computer scientist, Ph.D. from University of Texas at Austin under Risto Miikkulainen.
In his thesis covering machine learning and global optimization, he introduced the martingale based Evolutionary Annealing,
which resembles an evolutionary algorithms to approximates samples from an increasingly sharp
Boltzmann distribution, asymptotically focusing on the global optima.
Neuroannealing applies evolutionary annealing to evolve and learn neural networks.
His further research interests include humanoid robotics, deep learning and opponent modelling in games.
Selected Publications
- Alan J. Lockett, Charles L. Chen, Risto Miikkulainen (2007). Evolving Explicit Opponent Models for Game Play. GECCO 2007
- Alan J. Lockett, Risto Miikkulainen (2008). Evolving Opponent Models for Texas Hold 'Em. CIG 2008
2010 ...
- Alan J. Lockett, Risto Miikkulainen (2011). Real-Space Evolutionary Annealing. GECCO 2011
- Alan J. Lockett (2012). General-Purpose Optimization Through Information Maximization. Ph.D. thesis, University of Texas at Austin, advisor Risto Miikkulainen, pdf
- Alan J. Lockett, Risto Miikkulainen (2013). A Measure-Theoretic Analysis of Stochastic Optimization. FOGA 2013
- Alan J. Lockett (2014). Model-optimal optimization by solving bellman equations. GECCO 2014 [4]
- Alan J. Lockett, Risto Miikkulainen (2014). Evolutionary Annealing: Global Optimization in Arbitrary Measure Spaces. Journal of Global Optimization, Vol. 58
- Marijn F. Stollenga, Alan J. Lockett, Jürgen Schmidhuber (2015). The Natural Gradient as a control signal for a humanoid robot. Humanoids 2015
- Alan J. Lockett (2015). Insights From Adversarial Fitness Functions. FOGA 2015, pdf (extended abstract)