back to list

Project: Synthetic incomplete data generation

Description

Data generation is an important task, but typically the missing data mechanism is not fully modeled and exploited in the process. This project intends to study such a problem and to create tools for data generation with missing values. Besides data generation from random distributions, we will consider how to learn a model and generate data from it, which can be used for hiding the original data and for improving models by post-hoc fine tuning. We will focus on probabilistic generative machine learning models.

Details

Supervisor

Cassio de Campos

Interested?

Get in contact