This person can no longer be contacted through the School of Engineering and Computer Science at Victoria University of Wellington
Research Interests: Machine Learning, Neural Networks, Bayesian Optimization
In a great many domains, solving a problem involves a "mission critical" step in which there is a need to tune some set of parameters of the system to get the best possible outcome. Often background knowledge is at a premium and, as a result, this tuning has to be based predominantly on the results of previous experience with particular settings. How might one arrive at "optimal", or at least very good, settings without needing too much experience? Heuristics and guess-work may be sufficient, but when the expense of gaining that previous experience is high, the question of where best to try next becomes important enough to warrant theoretical investigation in its own right. The resulting field, known as Bayesian Optimization, has made heavy use of machine learning models, however, it has not been obvious how to use one in particular despite its general popularity, namely neural networks. I study novel methods using which neural networks can be leveraged in pursuit of the best samples when tuning a system.