DoubleMLDeep: Estimation of Causal Effects with Multimodal Data
This work addresses causal inference challenges for researchers and practitioners in fields like economics and medicine using non-traditional data, but it appears incremental as it adapts existing methods to multimodal contexts.
This paper tackles the problem of estimating causal effects using unstructured multimodal data (text and images) as confounders by proposing a neural network architecture adapted to the double machine learning framework and a semi-synthetic dataset for evaluation. The results highlight the potential benefit of using such data directly in causal studies, though no concrete performance numbers are provided.
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 and architectures are evaluated on the semi-synthetic dataset and compared to standard approaches, highlighting the potential benefit of using text and images directly in causal studies. Our findings have implications for researchers and practitioners in economics, marketing, finance, medicine and data science in general who are interested in estimating causal quantities using non-traditional data.