Student research opportunities
Machine Learning with Structured Representations in a Dataflow Framework
Project Code: CECS_665
This project is available at the following levels:
Honours, Masters
Keywords:
machine learning, software engineering
Supervisor:
Dr Stephen GouldOutline:
Many real-world machine learning problems follow a fairly standard pipeline of processing stages---feature extraction, parameter learning, evaluation, etc. Typically, features are encoded as vectors and a number of tools exist to facilitate this representation. However, many interesting machine learning tasks involve structured data. This project explores a dataflow framework to machine learning that allows for structure data (i.e., strings, trees, graphs, etc).
Requirements/Prerequisites
Students should have strong C/C++ programming experience and a background in machine learning and/or optimization.
