Faculty & Research

Samrat Roy


Samrat Roy is a faculty member at IIM Ahmedabad in the Operations and Decision Sciences Area. Prior to this position, Samrat worked as a postdoctoral researcher at the Wharton School (Department of Statistics and Data Science) in the University of Pennsylvania. He completed his M.STAT. from Indian Statistical Institute and received his Ph.D. in Statistics from the University of Florida (UF). Before starting his journey as a doctoral student at the UF, Samrat also worked at Ernst & Young and Credit Suisse as a credit risk model developer.

Throughout his academic and professional career, Samrat has traversed different paradigms of Statistics and Data Sciences including High-dimensional Statistics, Tensor Models, Causal Inference, Observational Studies, Time Series Models, and so on, with 10+ years of experience and training in cutting-edge technicalities.


Primary Area : Operations and Decision Sciences


Email : samratr@iima.ac.in

Phone : +91-79-7152 4930

Secretary : Pritesh Parmar

Phone : +91-79-7152 7973


Postdoctoral Researcher, Wharton School, University of Pennsylvania (Aug 2021 to Dec 2023). Primary advisors: Prof. Dylan Small and Prof. Sean Hennessy. Primary research area: Causal Inference and Observational Studies.

PhD in Statistics, University of Florida (Aug 2016 to Aug 2021). Thesis advisor: Prof. George Michailidis. Thesis topic: High Dimensional Tensor Models. Other research areas: Causal Inference, High Dimensional Time Series models.

M.STAT, Indian Statistical Institute (June 2012 to June 2014).

B.Sc. in Statistics, St. Xavier’s College Kolkata (June 2009 to June 2012).

Research Area

Causal Inference and Observational Studies.

High Dimensional Statistics – Tensor models.

Advanced Time Series Models.

Statistical Machine Learning – Advanced penalized estimation framework.

Professional Experience

Risk Model Developer, Credit Suisse Business Analytics, Mumbai (Apr 2015 to June 2016).

Credit Risk Consultant, Ernst Young LLP, Bangalore (June 2014 to Apr 2015).

Selected journal publications

  •  Roy, S., Michailidis, G. (2024). A Regularized Low Tubal-Rank Model for High-dimensional Time Series Data. Accepted in Statistica Sinica.
  •  Roy, S., Daniels, M., Roy, J. (2024). A Bayesian nonparametric approach for multiple mediators with applications in mental health studies. Published in Biostatistics.
  •  Roy, S., Ye, T., Ertefaie, A., Vo, T., Flory, J., Hennessy, J., Small, DS. (2023). Group Sequential Testing under Instrumented Difference-in-Differences approach. Published in Statistics in Medicine. 
  • Roy, S., Michailidis, G. (2022). Regularized High Dimension Low Tubal-Rank Tensor Regression. Published in Electronic Journal of Statistics.

Manuscripts submitted

  •  Roy, S., Bogomolov, M., Heller, R., Claridge, AM., Beeson, T., Small, DS. (2024). Exploratory Data Analysis, Confirmatory Data Analysis and Replication in the Same Observational Study: A Two Team Cross-Screening Approach to Studying the Effect of Unwanted Pregnancy on Mothers’ Later Life Outcomes.
  •  Vo. T., Ye, T., Ertefaie, A., Roy, S., Flory, J., Hennessy, J., Vansteelandt, S., Small, DS.(2024). Structural Mean Models for Instrumented Difference-in-Difference Design.

Selected manuscripts in progress

  • Chakraborty, N., Roy, S., Michailidis, G. (2024+). Low rank plus sparse panel VAR model with reduced dimensional data.
  • Paul, M., Roy, S., Small, DS. (2023+). A Weighted Inverse Normal method-based combination of U-statistics for matched pairs with missing data.
  • Roy, S., Chakraborty, N., (2024+). Low tubal-rank plus sparse dynamic model for matrix valued time series data.

Conference publications

  • Chakraborty, N., Roy, S., Leite, W., Faradonbeh, M. and Michailidis, G. (2021). The effects of a personalized recommendation system on students’ high-stakes achievement scores: A field experiment. Published in 14th International Conference on Educational Data Mining.
  • Leite, W., Roy, S., Chakraborty, N., Michailidis, G., Huggins-Manley, A., D’Mello, S., Faradonbeh, M., Kuang, H. and Jing, Z. (2021). An evaluation of a novel learning resource recommendation algorithm before and after school closures due to COVID-19. Published in 12th International Learning Analytics and Knowledge Conference.

Selected presentations as invited speaker

  • IISA Bangalore, 2022: Modern methods for high-dimensional and distributed multivariate analysis.
  • ISPE Student Chapter, 2022: Bootcamp on Instrumental variable (Invited Instructor)
  • ISBA 2022: Bayesian approaches for complex problems in causal inference
  • Research Day at UPenn DBEI, 2022
  • Center for Causal Inference, UPenn, 2022
  • CMStatistics, Berlin, 2023
  • IISA Cochin, 2024 (upcoming)

Awards and Scholarships

  • Best poster presentation award (non-student section), Conference on Advances in Bayesian and Frequentist Statistics, Rutgers University, April 2022.
  • NSF Junior Researcher Travel Grant, April 2022.
  • NSF (DMS-2150112) grant for ICSA, 2022
  • ISBA 2022 Junior Travel Support, sponsored by de Castro Statistics Initiative, Collegio Carlo Alberto.
  • NISS 2022 Junior Researcher Travel Grant
  • College of Liberal Arts and Sciences Graduate Travel Award, University of Florida, Aug 2019
  • Extra Miler Award, Earnst and Young , Aug 2014
  • INSPIRE scholarship, Ministry of Science Technology, Govt. of India, Jun 2009 - Jun 2010