SyGNet: Synthetic Data for the Social Sciences using Deep Learning

• In Event: Experimental Design and Analysis

Brief Overview

We explore how synthetic data can be used to assist researchers in experimental contexts, at both the design and analysis stages, and demonstrate how to produce these novel observations using generative adversarial networks, a form of deep learning.

Authors

• Thomas Samuel Robinson, London School of Economics and Political Science

• Maksim Zubok, University of Oxford

• Artem Nesterov, Durham University