Research Project Title:
Time series analysis with BayesDB
abstract:I will be contributing to BayesDB, an open-source AI platform developed by the group. Multivariate time-series data is increasingly available, but governments, non-profit organizations, and public policy initiatives currently struggle to find what is useful in it. The MIT Probabilistic Computing Project has developed an open-source AI platform, called BayesDB, that helps people more rapidly learn from data. MIT Probabilistic Computing Project has further developed a number of new time series modelling and inference approaches which are not yet integrated into the BayesDB platform. Those new capabilities have to potential to help domain experts to gain knowledge from time-series data and to automate parts of the data analysis process that previously were done by PhD-level human experts.
I am participating in SuperUROP as a way to tie together everything I've learned at MIT. I am excited to be able to apply my studies to something.