Research

Research interest:

• Computational Statistics

• Cluster Analysis

• Finite mixture modeling

• Hidden Markov models

• Change point estimation

Selected publications:

• Zhang, Y., Sarkar, S., Chen, Y., and Zhu, X., (2024) On regime changes in text data using hidden Markov model of contaminated vMF distribution. Data Mining and Knowledge Discovery (accepted).

• Asilkalkan, A., Zhu, X., and Sarkar, S., (2023) Tensor-variate time series modeling with hidden Markov models. Advances in Data Analysis and Classification, 1-18.

• Sarkar, S. and Zhu, X., (2022). Finite mixture model of hidden Markov regression with covariate dependence. Stat, 11(1), p.e469  (https://onlinelibrary.wiley.com/doi/full/10.1002/sta4.469)

  • Sarkar, S. and Zhu, X., (2022). Multiple change point clustering of count processes with application to California COVID data. Pattern Recognition Letters. (https://www.sciencedirect.com/science/article/pii/S0167865522000903?dgcid=author)

• Zhu, X., Sarkar, S, Melnykov, V., (2022). MatTransMix: An R Package for matrix parsimonious models, Journal of Classification, 39(1), pp.147-170. (preprint)

• Sarkar, S., Melnykov, V. and Zhu, X., (2021). Tensor-variate finite mixture modeling for the analysis of university professor remuneration. The Annals of Applied Statistics 15 (2), 1017-1036 (https://doi.org/10.1214/20-AOAS1420). (preprint)

• Melnykov, V., Sarkar, S. and Melnykov, Y., (2021). Finite mixture modeling of directed weighted multilayer networks, Pattern Recognition, 112, p.107641 (https://doi.org/10.1016/j.patcog.2020.107641). (preprint)

• Sarkar, S., Melnykov, V. and Zheng, R., (2020). Gaussian mixture modeling and model-based clustering under measurement inconsistency. Advances in Data Analysis and Classification, pp.1-35 (https://doi.org/10.1007/s11634-020-00393-9). (preprint)

• Sarkar, S., Zhu, X., Melnykov, V. and Ingrassia, S., (2020). On parsimonious models in matrix data mixture modeling. Computational Statistics and Data Analysis, 142, p.106822 (https://doi.org/10.1016/j.csda.2019.106822). (preprint)

Software development:

• Sarkar, S. and Melnykov, V. (2020) R-package netClust.

Selected talks:

• Asilkalkan, A., Zhu, X., and Sarkar, S. Finite mixture of hidden Markov models for tensor-variate time series data, Invited session talk (virtual) at 16th International Conference of the ERCIM WG on Computational and Methodological Statistics (CMStatistics), Berlin, Germany, Dec, 2023.

• Sarkar, S. and Zhu, X. Multiple change point clustering of count processes, Contributed session talk at Workshop on Model-Based Clustering and Classification, Catania, Italy, August, 2022.

• Sarkar, S. and Zhu, X. Multiple change point clustering of count processes, Invited session talk at 5th International Conference on Econometrics and Statistics (Hybrid EcoSta), Kyoto, Japan, June, 2022.

• Sarkar, S., Melnykov, V. and Zhu, X. Tensor-variate finite mixture model for the analysis of university professor remuneration, Invited session talk at CLAssification and Data Analysis Group (CLADAG) conference, Florence, Italy, September, 2021 (online due to covid-19 pandemic).

• Melnykov, V., Sarkar, S. and Melnykov, Y. Finite mixture modeling of directed weighted multilayer networks, Invited session talk at International Conference on Statistical Distributions and Applications (ICOSDA), Grand Rapids, MI, USA, October, 2019.

• Melnykov, V., Sarkar, S. and Melnykov, Y. Finite mixture modeling of directed weighted multilayer networks, Invited session talk at International Federation of Classification Societies (IFCS) conference, Thessaloniki, Greece, August, 2019.

• Sarkar, S., Zhu, X., Melnykov, V., and Ingrassia, S. Parsimonious models in matrix data mixture modeling, Lightning talk at Workshop on Model-Based Clustering and Classification, Catania, Italy, September, 2018.

• Sarkar, S., Melnykov, V., Zheng, R. and Zhu, X. On the use of transformations in finite mixture modeling, Invited departmental talk at University of Louisville, KY, March, 2018.

• Sarkar, S., Melnykov, V. and Zheng, R. Gaussian mixture modeling and model-based clustering under measurement uncertainty, Joint Statistical Meeting, Baltimore, MD, August, 2017.

Student advising:

• Graduate level:

  • Josiah Leinbach, Master’s thesis - Authorship attribution in ancient text (tentative title). (2024 - adviser and thesis committee chair)

  • Lada Carlisle, Project for M.S. in Data Science - Studying regime change in Academy Award best pictures. (2023 - faculty supervisor)

  • Kehinde Fagbamigbe, Master’s thesis - Examining gender equality in Unite States undergraduate enrollment using hidden Markov model. (2023 - adviser and thesis committee chair)

  • Justin Peter, Project for M.S. in Data Science - Application of von Mises-Fisher mixture model for text clustering IMDB movie synopses. (2022 - faculty supervisor)

  • Shabnam Kian Khah, Master’s thesis - County level clustering of US COVID-19 cases and mortality using matrix mixture model. (2021 - adviser and thesis committee chair)

  • Ismail Olayemi, Master’s thesis - Clustering COVID-19 cases in the United States by county and socio economic characteristics (2021 - adviser and thesis committee chair)

• Undergraduate level:

  • Emily Eskuri, Honors project - An Exploratory Analysis of the BGSU Learning Commons Student Usage Data (2021 - co-adviser)

  • Chaska Noel McGowan, Capstone project - Statistical Analysis of Land Use Conversion Trends in Northwest Ohio (2020 - co-adviser)