07-10, 15:25–15:55 (US/Pacific), Room 316
Causal inference has traditionally been used in fields such as economics, health studies, and social sciences. In recent years, algorithms combining causal inference and machine learning have been a hot topic. Libraries like EconML and CausalML, for instance, are good Python tools that facilitate the easy execution of causal analysis in areas like economics, human behavior, and marketing. In this talk, I will explain key concepts of causal inference with machine learning, show practical examples, and offer some practical tips. Attendees will learn how to apply machine learning to causal analysis effectively, boosting their research and decision-making.
The integration of causal inference with machine learning represents a significant advancement in understanding cause-and-effect relationships across various fields, such as economics, healthcare, and marketing. This presentation aims to clarify the concepts of causal inference with machine learning, offering insights into its practical applications, challenges, and solutions.
Slide
https://docs.google.com/presentation/d/1XFM0j_D_5Qa1S3lEkTQcVKO4ONwOY5GchZ4KImgl-5Y/edit?usp=sharing
Demo code
https://github.com/takechanman1228/Effective-Uplift-Modeling
Agenda :
What is Causal Inference?
- Explanation of traditional causal inference
- Discussion on Randomized Controlled Trials (RCT), selection bias, confounders, counterfactuals, and Simpson's Paradox
What is Causal Inference with Machine Learning?
- History and evolution of causal inference with machine learning
- Key differences between traditional machine learning and causal machine learning
Technique #1 : Meta Learners
- Introduction to the Potential Outcomes Framework and Treatment Effects (ATE, CATE, ITE)
- Overview of Meta Learners methods
Case Study in Economics
- Case study based on the National Supported Work (NSW) social job training program
- Demonstration of framing, estimation, and refutation processes using EconML to measure ATE
Technique #2 : Uplift Modeling
- Purpose of uplift modeling, Customer segmentation for targeting
- Methods for uplift modeling (Meta-learner, Decision Tree-Based Method)
- Explanation of the Uplift Tree algorithm
Case Study in Marketing
- Case study based on the Criteo Dataset for ad targeting
- Demonstration of training and evaluating uplift models using CausalML
- Explanation of evaluation methods (Uplift Curve, AUUC score)
Key Takeaways:
- Gain a comprehensive understanding of the key concepts and primary methods used in combining Causal Inference with Machine Learning.
- Learn about the latest tools and libraries such as EconML and CausalML.
- Discover practical examples of how causal inference can be applied to real-world scenarios.
Target Audience:
- Researchers, academics, and students in economics, health studies, social sciences, and data science, particularly those working with observed data
- Data scientists, analysts, and developers interested in enhancing their business practices, especially in marketing, retail, e-commerce, and customer analytics areas
Hajime is a data professional with five years of expertise in marketing, retail, and eCommerce, working across Japan and the United States.
As a Data Analyst at Procter and Gamble and MIKI HOUSE Americas, Hajime has led data-driven strategy formulation and implemented technology initiatives such as e-commerce expansion, advertising optimization, and the identification of growth opportunities.