Physics Quantities: While I have much experience with quantities derived from sparse optimization methods I’m really open to anything involving solid state physics and optimization.

Learned Dictionaries: I’m interested in methods that learn dictionaries from a set of quantities of a certain type, say \(Q_i\). In the case of sparse coding, these dictionaries are learned so that any \(Q_i\) can be described as a sparse linear combination of the dictionary functions.

Classification: General classification of physics quantities to build tools for scientists. I’m also interested in methods that utilize our ability to define \(Q\) in terms of \(D\) so that we may build accurate models to classify \(Q\).

Evolving Methods: Methods that allow us to utilize learned dictionaries within the calculation of future quantities to produce better methods. A feedback loop that uses prior solutions to guide future solutions.