LiberTEM : Distributed algorithms for dimensionality reduction methods on scanning transmission electron microscopy (STEM) data

Dimensionality reduction techniques are useful methods that allow us to gain crucial insights about the given dataset. Unfortunately, such methods become computationally intensive when dealing with large scale dataset. To deal with complexity issues, one possible approach is to implement algorithms in a distributed fashion. Ideally, users of LiberTEM can benefit from implementation of these algorithms that they can run through a simple pipeline called User-defined functions, which allow the users to run functions with their desired functionality without having to worry about parallelization, which is done under the hood by LiberTEM. Therefore, my project will be concerning both the distributed implementation of a dimensionality reduction method as well as improving on the User-defined functions framework.