Skip to content

Latest commit

 

History

History
139 lines (94 loc) · 5.02 KB

workshop.md

File metadata and controls

139 lines (94 loc) · 5.02 KB

Workshop: Causal Modeling in Machine Learning

Instructor: Robert Osazuwa Ness
Author of Causal AI

Overview

This full-day workshop introduces participants to the foundations and modern applications of causal modeling in machine learning. We focus on practical, model-based reasoning using interventions, causal graphs, and counterfactuals — culminating in how these tools integrate with contemporary machine learning, including deep generative and probabilistic models. This workshop is a companion to Probabilistic Machine Learning Workshop.

The workshop is designed for practitioners and researchers who want to build ML systems that reason robustly about cause and effect, go beyond correlations, and support decision-making in complex real-world environments.

Learning Objectives

By the end of this workshop, participants will be able to:

  • Construct causal graphs and reason about dependencies, interventions, and confounding.
  • Apply do-calculus to determine identifiability of causal effects.
  • Estimate causal effects using regression adjustment, propensity methods, and modern ML estimators.
  • Perform counterfactual reasoning using structural causal models and algorithmic counterfactuals.
  • Integrate causal reasoning with A/B testing, bandits, reinforcement learning, and sequential decision-making.
  • Understand how deep latent variable models and probabilistic programming support counterfactual inference.

Target Audience

This workshop is well suited for:

  • ML researchers and data scientists
  • Experimentation platform teams
  • Applied scientists in tech, healthcare, finance, social science, or policy
  • PhD students and academics studying causal inference or causal ML
  • Engineers integrating causal reasoning into products, decision systems, or simulations

Participants should be comfortable with probability, random variables, and basic machine learning.

Schedule (Adjustable)

8:00–8:30 — Introduction & Motivation

  • Why causality is essential in modern ML
  • Examples where predictive accuracy fails but causal reasoning succeeds
  • Causal ML in industry: experimentation, personalization, simulation, safety-critical systems

8:30–9:45 — Causal Graphs & Structural Models

  • Directed acyclic graphs (DAGs)
  • Causal vs. statistical relationships
  • Structural causal models (SCMs)
  • Conditional independence, d-separation, and graphical reasoning
  • Case study: confounding and backdoor criteria

9:45–10:00 — Break

10:00–11:30 — Interventions & Causal Inference

  • The do-operator and intervention semantics
  • Identification of causal effects
  • Adjustment sets and backdoor/frontdoor criteria
  • Matching, weighting, and ML-based estimation
  • Hands-on examples and intuition-building graphical exercises

11:30–12:30 — Lunch Break

12:30–13:45 — Counterfactuals & Algorithmic Counterfactuals

  • Counterfactual semantics in SCMs
  • Causal queries vs. statistical predictions
  • Individual-level vs. population-level counterfactuals
  • Algorithmic counterfactuals in ML systems
  • Applications in fairness, reasoning, simulation, and generative models

13:45–14:00 — Break

14:00–15:30 — Causal Modeling in ML Systems

  • Causal inference with machine learning estimators
  • Causal forests, meta-learners (T-, S-, X-, R-learners)
  • Causal representation learning
  • Causal reasoning in reinforcement learning and decision-making systems
  • Case study: experimentation and adaptive policies

15:30–15:45 — Break

15:45–17:00 — Deep Causal Models & Future Directions

  • Deep causal latent variable models
  • Causal generative models and structural VAEs
  • Counterfactual simulation using probabilistic programming
  • Causality for world models, safety, and alignment
  • A practitioner’s playbook:
    • How to choose the right causal tools
    • How causal reasoning complements ML
    • Pitfalls and anti-patterns
    • Recommended reading and research roadmap
  • Open Q&A

Software and Tools

While the workshop is conceptual and model-based, demonstrations use:

  • Pyro for probabilistic programming and counterfactual simulation
  • Lightweight PyTorch examples for causal estimators and structural models

(Hands-on notebooks may be provided depending on event format.)

Suggested Background Reading

  • Causal AI — Robert Osazuwa Ness
  • Causal Inference in Statistics: A Primer — Pearl, Glymour, Jewell
  • Causality — Judea Pearl
  • Key tutorials on SCMs, do-calculus, and ML-based causal inference

Instructor

Robert Osazuwa Ness is a researcher specializing in causal AI, generative modeling, probabilistic programming, and counterfactual reasoning. He has worked as an AI research scientist in both industry and academia and is the author of Causal AI.