My research develops theories and algorithms for causal inference with the goal of personalized, data-driven decision-making. Specifically, I build methods to answer individual-level rather than population-level and distributional rather than mean causal questions by using tools and frameworks from statistics and machine learning, such as exponential family and matrix completion. I also develop fair and private algorithms for supervised machine learning.

Here, I categorize my work into the following research topics:

• Causal Inference
• Exponential Family
• Fairness and Privacy

For a chronological list, please refer to my CV. Below, † denote alphabetical ordering and a superscript * denotes equal contribution.

Causal Inference

• Doubly robust inference for causal latent factor models.
  Alberto Abadie†, Anish Agarwal, Raaz Dwivedi, Abhin Shah
  Under review
  Arxiv

• On counterfactual inference with unobserved confounding.
  Abhin Shah, Raaz Dwivedi, Devavrat Shah, Gregory W. Wornell
  Under review
  Arxiv Slides Poster

• Front-door adjustment beyond Markov equivalence with limited graph knowledge.
  Abhin Shah, Karthikeyan Shanmugam, Murat Kocaoglu
  37th Conference on Neural Information Processing Systems (NeurIPS), 2023
  Arxiv Poster

• Finding Valid Adjustments under Non-ignorability with Minimal DAG Knowledge.
  Abhin Shah, Karthikeyan Shanmugam, Kartik Ahuja
  25th International Conference on Artificial Intelligence and Statistics (AISTATS), 2022
  Paper Slides Poster Video Code

• Treatment effect estimation using invariant risk minimization.
  Abhin Shah, Kartik Ahuja, Karthikeyan Shanmugam, Dennis Wei, Kush Varshney, Amit Dhurandhar
  IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021
  Paper Slides Poster Video Code

Exponential Family

• On Computationally Efficient Learning of Exponential Family Distributions.
  Abhin Shah, Devavrat Shah, Gregory W. Wornell
  IEEE Transactions on Information Theory, 2024
  Arxiv

• A Computationally Efficient Method for Learning Exponential Family Distributions.
  Abhin Shah, Devavrat Shah, Gregory W. Wornell
  35th Conference on Neural Information Processing Systems (NeurIPS), 2021
  Paper Slides Poster Video

• On learning Continuous Markov Random Fields.
  Abhin Shah, Devavrat Shah, Gregory W. Wornell
  Oral at 24th International Conference on Artificial Intelligence and Statistics (AISTATS), 2021
  Paper Slides Poster Video

Fairness and Privacy

• Group Fairness with Uncertainty in Sensitive Attributes.
  Abhin Shah, Maohao Shen, Jongha Jon Ryu, Subhro Das, Prasanna Sattigeri, Yuheng Bu, Gregory W. Wornell
  IEEE International Symposium on Information Theory (ISIT), 2024
  Arxiv Slides Poster

• Selective Regression Under Fairness Criteria.
  Abhin Shah*, Yuheng Bu*, Joshua Ka-Wing Lee, Subhro Das, Rameswar Panda, Prasanna Sattigeri, Gregory W. Wornell
  39th International Conference on Machine Learning (ICML), 2022
  Paper Slides Poster Video Code News

• Optimal Compression of Locally Differentially Private Mechanisms.
  Abhin Shah, Wei-Ning Chen, Lucas Theis, Peter Kairouz, Johannes Ballé
  25th International Conference on Artificial Intelligence and Statistics (AISTATS), 2022
  Paper Slides Poster Video Code