DoubleMLDeep: Estimation of Causal Effects with Multimodal Data

This paper explores the use of unstructured, multimodal data, namely text and images, in causal inference and treatment effect estimation. We propose a neural network architecture that is adapted to the double machine learning (DML) framework, specifically the partially linear model. An additional contribution of our paper is a new method to generate a semi-synthetic dataset which can be used to evaluate the performance of causal effect estimation in the presence of text and images as confounders. The proposed methods …

architecture causal inference cs.ai cs.lg data econ.em effects framework images inference linear linear model machine machine learning multimodal multimodal data network network architecture neural network paper stat.me stat.ml text treatment unstructured