Avatar

Amit Sharma

Senior Researcher

Microsoft Research India

Biography

2019.05.30: Using counterfactual examples to explain machine learning. [Paper and Code].

2018.08.19: Emre and I gave a tutorial on causal inference at KDD. [Slides]

2018.08.01: Released a Python library for causal inference, DoWhy. [Blog][Github]


Data tells stories. My research aims to tell the causal story.

As machine learning systems move into societally critical domains such as healthcare, education, finance and criminal justice, questions on their impact gain fundamental importance. The key insight in my work is to consider modern algorithms as interventions, just like a medical treatment or an economic policy. Unlike typical interventions studied in social and biomedical sciences, however, algorithmic interventions can be arbitrarily complex. I work on developing methods to estimate causal impact of such systems and build algorithms that optimize the causal effect. I am also passionate about designing new interventions for societal impact, especially in healthcare.

If you are interested in working with me at MSR India, drop me an email. We hire interns throughout the year. There are also postdoctoral positions available. Additionally, if you are an undergraduate or a masters student, our lab runs an excellent pre-doctoral Research Fellowship program.

Read more…

Interests

  • Causal Inference
  • Computational Social Science
  • Machine Learning & Societal Impact

Education

  • PhD in Computer Science, 2015

    Cornell University

  • B.Tech. in Computer Science, 2010

    IIT Kharagpur

Recent Talks

The impact of computing systems | Causal inference in practice

Computing and machine learning systems are affecting almost all parts of our lives and the society at large. How do we formulate and …

Measuring Effectiveness of ML Systems

Recent Posts

Trip report from ACM COMPASS: 2nd Conference on Computing and Sustainable Societies

Last year, I attended the inaugural ACM conference on Computing and Sustainable Societies (COMPASS) and was immediately sold on the …

A Gentle Introduction to Causal Inference

That we find out the cause of this effect, Or rather say, the cause of this defect, For this effect defective comes by cause. …

A Simple Guide to Doubly Robust Estimation

Two roads diverged in a wood, and I— I took the one less traveled by, But if I could go back, I will try To take both: why …

Cumulative Distribution Plots for Frequency Data in R

R has some great tools for generating and plotting cumulative distribution functions. However, they are suited for raw data, not when …

Contact