skills.rest
sheehe
sheeheCOMMUNITY
Explore all skills by this author →
GitHub

ml-causal

Estimate ATE and CATE from observational data using DML and causal forests.

Updated Apr 15, 2026
Search Tags:#dml#machine-learning#causal-inference#blp#heterogeneous-effects#causal-forest#aipw
One-click install
npx skills add https://github.com/sheehe/coase --skill ml-causal
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ml-causal
Source: https://github.com/sheehe/coase/tree/main/%E5%AE%9E%E8%AF%81%E7%A7%91%E7%A0%94%E6%8F%92%E4%BB%B6/econometrics/econometrics/skills/ml-causal
Command: npx skills add https://github.com/sheehe/coase --skill ml-causal

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Estimates causal effects from observational data using ML-based nuisance estimation to combine flexible modeling with rigorous identification.

Core Features & Use Cases

  • Double/Debiased Machine Learning (DML) for partially linear models to estimate the causal parameter θ and ATE.
  • Causal Forest (GRF) for estimating heterogeneous treatment effects (CATE) and exploring effect heterogeneity.
  • BLP (Best Linear Predictor) and CLAN analyses to assess variation in CATE across observables.
  • AIPW / DR learners and meta-learners for robust CATE estimation with cross-fitting.

Quick Start

Prepare your data (Y, D, X) and select a method (e.g., DML or causal forest) to estimate treatment effects and CATE.

Frequently Asked Questions about ml-causal

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I estimate heterogeneous treatment effects from observational data using machine learning?▼

Estimate heterogeneous treatment effects by applying ML-based nuisance estimation to observational data. Techniques like causal forest and DR learners isolate the CATE by flexibly modeling nuisance functions, producing robust effect estimates across observable subgroups.

What is Double/Debiased Machine Learning (DML) for causal inference?▼

Double/Debiased Machine Learning (DML) is a causal inference technique for partially linear models that estimates the causal parameter and ATE. It uses cross-fitting and ML models for nuisance functions to remove bias, yielding robust causal estimates from observational data.

Can I use AIPW and meta-learners for CATE estimation with cross-fitting?▼

Yes, you can use AIPW and meta-learners for robust CATE estimation. These methods rely on cross-fitting to estimate nuisance functions, combining flexible ML modeling with rigorous identification strategies to produce valid heterogeneous effect estimates.

How do I assess variation in conditional average treatment effects across observables?▼

Assess variation in conditional average treatment effects using BLP (Best Linear Predictor) and CLAN analyses. These methods evaluate effect heterogeneity by probing the relationship between estimated CATEs and observable covariates after applying causal forest.

Do I need a valid identification strategy like an IV or RCT to use causal forest?▼

Yes, you need a valid identification strategy such as an RCT, IV, or DID where appropriate. Causal forest and DML require cross-fitting and robust assumptions to combine ML nuisance estimation with valid causal inference successfully.

Related Skills

Explore More Data & Analytics Skills→
James-TrainaJames-Traina
11

causal-inference

Implement causal identification strategies like IV/2SLS, DiD, RDD, synthetic control, and matching.

Community
Advanced
zhouziyue233zhouziyue233
6

ml-causal

Automate ML-based causal inference to estimate treatment effects in econometrics.

Community
Advanced
xjtulycxjtulyc
33

ml-econometrics

Estimate causal treatment effects with Post-LASSO and DML cross-fitting.

Community
Advanced
nahisahonahisaho
3

scientific-causal-ml

Estimate heterogeneous treatment effects with DoWhy, EconML, and Double ML.

Community
Advanced