Transfer Learning for Estimating Causal Effects using Neural Networks
arXiv:1808.07804
Abstract
We develop new algorithms for estimating heterogeneous treatment effects, combining recent developments in transfer learning for neural networks with insights from the causal inference literature. By taking advantage of transfer learning, we are able to efficiently use different data sources that are related to the same underlying causal mechanisms. We compare our algorithms with those in the extant literature using extensive simulation studies based on large-scale voter persuasion experiments and the MNIST database. Our methods can perform an order of magnitude better than existing benchmarks while using a fraction of the data.
References in corpus (7)
- Prototypical Networks for Few-shot Learning
- DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition
- Recursive Partitioning for Heterogeneous Causal Effects
- Theoretical Models of Learning to Learn
- Meta-learners for Estimating Heterogeneous Treatment Effects using Machine Learning
- Matching Networks for One Shot Learning
- Overlap in Observational Studies with High-Dimensional Covariates
Cited by in corpus (6)
- Uplift Modeling for Multiple Treatments with Cost Optimization
- Everything Has a Cause: Leveraging Causal Inference in Legal Text Analysis
- A unified survey of treatment effect heterogeneity modeling and uplift modeling
- Balance Regularized Neural Network Models for Causal Effect Estimation
- Assessment of Heterogeneous Treatment Effect Estimation Accuracy via Matching
- Heterogeneous Treatment Effect Estimation through Deep Learning