Review_of_Deep_Learning_Methods_for_Individual_Treatment_Effect_Estimation_with_Automatic_Hyperparameter_Optimisation.pdf (332.62 kB)
Download file

Review of Deep Learning Methods for Individual Treatment Effect Estimation with Automatic Hyperparameter Optimization

Download (332.62 kB)
posted on 2022-12-06, 04:50 authored by Andrei SirazitdinovAndrei Sirazitdinov, Marcus BuchwaldMarcus Buchwald, Jürgen HesserJürgen Hesser, Vincent Heuveline

Abstract—Estimation of individual treatment effect (ITE) for different types of treatment is a common challenge in therapy assessments, clinical trials and diagnosis. Deep learning methods, namely representation based, adversarial, and variational, have shown promising potential in ITE estimation. However, it was unclear whether the hyperparameters of the originally proposed methods were well optimized for different benchmark datasets. To solve these problems, we created a public code library containing representation-based, adversarial, and variational methods written in TensorFlow. In order to have a broader collection of ITE estimation methods, we have also included neural network based meta-learners. The code library is made accessible for reproducibility and facilitating future works in the field of causal inference. Our results demonstrate that performance of most methods can be improved using automatic hyperparameter optimization. Additionally, we review the methods and compare the performance of the optimized models from our library on publicly available datasets. The potential of hyperparameter optimization may encourage researchers to focus on this aspect when creating new methods for inferring individual treatment effect.


LeMeDaRT - Digital Progress Hub for Health: Lean Medical Data - the right data at the right time. Support of patient journeys from prevention to top-quality care in tertiary care centers

Federal Ministry of Education and Research

Find out more...

'PERPAIN - Improvement of the treatment results of chronic musculoskeletal pain disorders through a personalized therapy approach'

Federal Ministry of Education and Research

Find out more...


Email Address of Submitting Author

ORCID of Submitting Author


Submitting Author's Institution

Mannheim Institute for Intelligent Systems in Medicine (MIISM), Engineering Mathematics and Computing Lab (EMCL), Interdisciplinary Center for Scientific Computing (IWR) and Heidelberg Institute for Theoretical Studies (HITS)

Submitting Author's Country

  • Germany