Topics studied include:
1. Evolutionary Annealing -- a martingale-based approach to optimization.
2. Neuroannealing -- using evolutionary annealing to learn neural networks.
3. Formal analysis of Iterative Stochastic Optimizers -- a measure theoretic approach to analyzing optimizer performance.
I completed the Ph.D. in May 2012 with thesis "General-Purpose Optimization through Information Maximization".
In May 2012, I was awarded a $180K NSF grant under the International Research Fellows Program. I will be studying as a postdoctoral researcher with Prof. Juergen Schmidhuber at IDSIA in Lugano, Switzerland beginning in June 2013. The topic for the research is "Deep Neural Networks for the Integration of Perception and Action in Robotic Controllers".
alan lockett [at] gmail com