teaching
lectures, tutorials, research advising
Tutorials on causal inference and machine learning
[Beginner]
If you are new to causal inference, try this tutorial first.
📖A breezy introduction to causal inference, involving fire, Hooke’s law and R. A. Fisher. [IC2S2] [Video]
[Effect inference]
If you are interested in effect inference and want to understand which method is suitable for your problem, this tutorial is a timeless one (except the early description of DoWhy!).
📖Part I: Learn about the two main frameworks (Causal Graphical Models and Potential Outcomes) and how they can be used together. [KDD 2018: Intro]
📖Part II: Learn about the two main methods (conditioning-based and natural experiments) and when to use which. [KDD 2018: Methods]
📖Part III: Case studies on causal inference with large-scale data [KDD 2018: Large Scale Data]
[For practitioners]
If you are a practitioner and interested in applying causal inference to your problem, I recommend this webinar.
📖Learn about the four steps of causal inference (Model, Identify, Estimate, Refute) and how to apply them using DoWhy/PyWhy. [MSR Webinar 2020] [Video]
💻Hands-on tutorial on causal inference and its connections to ML using DoWhy and EconML libraries. [PyWhy]
[Causality for ML]
If you are interested in how causality can improve robustness or generalization of machine learning, this tutorial explains what is possible.
📖Part I: When can causality help ML models generalize better, and when is it less useful. How to build trustworthy and causal ML systems? [MLRS 2023: Intro]
📖Part II: Learn about how large language models may address the biggest impediment to scaling causal ML: building a causal graph. [MLRS 2023: LLMs]
Causal reasoning book
A few early chapters on a forthcoming book with Emre Kiciman.
Book Outline: Causal Reasoning: Fundamentals and Machine Learning Applications
Research advising
Former students with their first positions after graduating.
MSR Research Fellows (1-2 years pre-doctoral programme for undergraduate students at MSR India)
- Sachin Pendse → Georgia Tech. Ph.D.
- Divyat Mahajan → MILA Ph.D.
- Ramaravind Mothilal → Univ. of Toronto Ph.D.
- Saloni Dash → Univ. of Washington Ph.D.
- Abhinav Kumar → MIT Ph.D.
- Jivat Kaur → UC Berkeley Ph.D.
- Parikshit Bansal → UT Austin Ph.D.
- Aniket Vashishtha → UIUC M.S.
Ph.D. students (co-supervised)
- Sachin Pendse, Georgia Tech. → Asst. Professor UCSF
- Gowtham Abbavaram, IITH → Postdoc @ CISPA Helmholtz Center, Germany