Data

Tools

Indexes

Applications

Experiments

Open Source Science

learning-representations-causal-inference

Code supplement for "Neuroevolutionary representations for learning heterogeneous treatment effects"

https://github.com/burkh4rt/learning-representations-causal-inference

Science Score: 67.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ✓
    DOI references
    Found 3 DOI reference(s) in README
  • ✓
    Academic publication links
    Links to: zenodo.org
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (9.6%) to scientific vocabulary

Keywords

causal-inference conditional-average-treatment-effect counterfactual-inference feature-engineering genetic-algorithms heterogeneous-treatment-effects learning-representations neuroevolution
Last synced: 6 months ago · JSON representation ·

Repository

Code supplement for "Neuroevolutionary representations for learning heterogeneous treatment effects"

Basic Info
Statistics
  • Stars: 1
  • Watchers: 1
  • Forks: 1
  • Open Issues: 0
  • Releases: 2
Topics
causal-inference conditional-average-treatment-effect counterfactual-inference feature-engineering genetic-algorithms heterogeneous-treatment-effects learning-representations neuroevolution
Created almost 2 years ago · Last pushed almost 2 years ago
Metadata Files
Readme Citation

README.md

DOI

Neuroevolutionary representations for learning heterogeneous treatment effects

Within the field of causal inference, we consider the problem of estimating heterogeneous treatment effects from data. We propose and validate a novel approach for learning feature representations to aid the estimation of the conditional average treatment effect or CATE. Our method focuses on an intermediate layer in a neural network trained to predict the outcome from the features. In contrast to previous approaches that encourage the distribution of representations to be treatment-invariant, we leverage a genetic algorithm that optimizes over representations useful for predicting the outcome to select those less useful for predicting the treatment. This allows us to retain information within the features useful for predicting the outcome even if that information may be related to treatment assignment. We validate our method on synthetic examples and illustrate its use on a real life dataset.

This code accompanies our paper[^1] and is broken into folders:

  • casual_framework/ defines a causal dataset structure and causal metalearners that use standard regression methods to perform inference, along with our machinery for feature engineering

  • generated_examples/ reproduces numerical experiments described in the paper on synthetic data

The code can be run in a conda environment —

sh conda env create -f environment.yml conda activate pearl python3 experiments_synthetic/nie_wager.py

[^1]: M. Burkhart and G. Ruiz, Neuroevolutionary representations for learning heterogeneous treatment effects, Journal of Computational Science 71 (2023)

Owner

  • Name: Michael Burkhart
  • Login: burkh4rt
  • Kind: user
  • Company: University of Cambridge

research associate—machine learning for neuroscience

Citation (CITATION.cff) 

cff-version: 1.2.0
message: "Please cite the following work when using this software."
license:
  - MIT
preferred-citation:
  type: article
  authors:
    - family-names: "Burkhart"
      given-names: "Michael C."
      orcid: "https://orcid.org/0000-0002-2772-5840"
    - family-names: "Ruiz"
      given-names: "Gabriel"
      orcid: "https://orcid.org/0000-0003-3349-760X"
  doi: "10.1016/j.jocs.2023.102054"
  journal: "Journal of Computational Science"
  title:
    "Neuroevolutionary representations for learning heterogeneous treatment
    effects"
  volume: 71
  year: 2023

GitHub Events

Total
  • Watch event: 1
Last Year
  • Watch event: 1

Dependencies

environment.yml pypi
pyproject.toml pypi