Research Project Title:
Applications of Learned Data Structures
abstract:MIT's Data Science and AI Lab (DSAIL) has shown that optimizing data structures with machine learning can have significant performance benefits over their generic counterparts in databases. These optimized models are known as learned structures. The goal of my project is to evaluate existing learned structures outside of a database context and explore the scope in which other C/C++ structures can benefit from machine learning optimization. Based on my findings, I would also like to build a general purpose learned structures toolkit for developers such that others can easily harness the benefits of DSAIL's learned structures.
"SuperUROP is my chance to do advanced research and gain expertise in a practical machine learning application. I hope to publish a paper and continue my research as an MEng student. My previous experience includes advanced subjects in ML, being a lab assistant in Course 6.036 (Introduction to Machine Learning), and three years of professional software development. I can't wait to test my hypotheses, and I have high hopes that I can substantially impact modern c++ programs."