Causality Seminar Series by Karthikeyan VS at IISc Bangalore
IISc Bangalore brings to you a seminar series by Karthikeyan V Shanmugam from Google Research. As an expert on the field of causal discovery and inference, with several key results in this area, Karthikeyan brings several new perspectives to our understanding of causality. This series of lectures will cover various aspects of causal discovery and causal inference.
Karthikeyan Shanmugam is a Research Scientist at Google Research India (Bengaluru) since April 2022. He is a part of the Machine Learning Foundations and Optimization Team. Previously, he was a Member of Research Staff with IBM Research AI, NY and a Herman Goldstine Postdoctoral Fellow at IBM Research, NY. He obtained his Ph.D. from UT Austin where his advisor was Alex Dimakis. He obtained his master’s degree from the University of Southern California, and prior to that his B.Tech (Dual Degree) from IIT Madras. His research interests broadly lie in Graph algorithms, Machine learning, Optimization, Coding Theory, and Information Theory. Specifically in machine learning, my recent focus is on Causal Inference, Bandits/RL and Explainable AI. All are invited. Snacks and Tea will be served post the talks.
Please join us on Mondays and Thursdays at the CSA Lecture Hall (Room 254), starting at 04.00pm. We would also be hosting the sessions online for those who are unable to join us in person.
Schedule
September 21 (Thursday)
Download Calendar: Thursday, 21st September 2023 16:00-17:30
In the First Lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Sure thing principle and its causal variant.
  2. Simpson's paradox - main motivating central issue in causal inference.
  3. Potential Outcomes Model - Formal Setting.
  4. Confounder balancing under Ignorability.
  5. Rubin Rosenbaum results on importance weighing and stratification.
September 25 (Monday)
Download Calendar: Monday, 25th September 2023 16:00-17:30
In the second lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Various estimators for ATE (Average Treatment Effect) under ignorability.
  2. CATE estimation and estimators for it.
  3. Individual treatment effect estimators – X learner and T learner.
  4. Representation learning for ITE – CFR Net, TAR net.
October 05 (Thursday)
Download Calendar: Thursday, 05th October 2023 16:00-17:30
In the third lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Continue representation learning for treatment effect.
  2. Introduction to Sensitivity Analysis.
  3. Doubly Robust estimator and its analysis.
  4. Instrumental Variable approach for linear models.
  5. Discuss open problems in these areas.
October 19 (Thursday)
Download Calendar: Thursday, 19th October 2023 16:00-17:30
In the fourth lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Individual Treatment Effect.
  2. X-learner S-learner and T-learner.
  3. Sample complexity analysis of these estimators.
October 26 (Thursday)
Download Calendar: Thursday, 26th October 2023 16:00-17:30
In the fifth lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Continuation of TarNets.
  2. Matching and Ignorability under transformations
  3. Instrumental Variables
October 30 (Monday)
Download Calendar: Thursday, 26th October 2023 16:00-17:30
In the sixth lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Looking back at Simpson’s paradox – what is ignorability does not hold.
  2. Introduction to Pearlian Approach to Causality.
  3. Definition of Structural Equation Models, Causal Bayesian Networks.
  4. Do calculus definitions.
  5. Equivalence to the modularity or interventional invariance condition.
November 02 (Thursday)
Download Calendar: Thursday, 02nd November 2023 16:00-17:30
In the seventh lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Proof of d-separation being equivalent to the local Markov condition
  2. Definition of Structural Causal Model.
  3. Proof that interventional distributions can be obtained from joint distribution
November 06 (Thursday)
Download Calendar: Monday, 06th November 2023 16:00-17:30
In the eighth lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Do calculus rules from d separation over augmented graphs
  2. Back door adjustment
  3. Tian and Pearl's do-separation over augmented graphs
November 16 (Thursday)
Download Calendar: Thursday, 16th November 2023 16:00-17:30
In the ninth lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. Continuation of front door criteria
  2. Examples of front door and back door adjustment in graphs
  3. Aspects of the ID Algorithm
November 30 (Thursday)
Download Calendar: Thursday, 30th November 2023 16:00-17:30
In the tenth lecture, Dr. Karthikeyan Shanmugam will cover the following topics:
  1. c-component factorization
  2. ID Algorithm
Seminar Series Details


Dates: Mondays and Thursdays, from 21st September 2023.
Location: CSA Seminar Hall (Room 254)
Affiliation: Google Research India