Biostatistics and Data Science

RESEARCH INTEREST

Application of Bayesian Statistics in Biomedicine, Clinical trials, Network Meta-Analysis, Bootstrap, Observed Confidence Levels, Analysis of Network Graph Models, and Survival Analysis.

Dr. Ghosh’s primary research interests lie broadly in statistical methodology. During his dissertation, he worked on the application of the bootstrap to complex statistical problems such as constructing multivariate confidence regions and computing observed confidence levels. Dr. Ghosh is currently working on developing Bayesian methodology for clinical trials, such as non-inferiority trials, enrichment trials. Alongside, he is expanding his research interests to include Meta Analysis, Network Data Analysis, and Survival Analysis. He also enjoys collaborative research with scholars from other disciplines. 

TEACHING AREAS

Mathematical Statistics, Biostatistics, Bootstrap, Stochastic Processes, Statistical Methods of Forecasting.

SELECTED PUBLICATIONS

• Ghosh, Santu,Chatterjee Arpita and Balakrishnan, N. (2017+): Nonparametric Confidence Intervals for Ranked Set Samples. Computational Statistics In Press , DOI: 10.1007/s00180-017-0744-0.
• Sawyer, Alexandra, Wise, Linda, Ghosh, Santu, Bhatia, Jatinder, and Stansfield, Brian (2017+): Comparison of Transfusion Thresholds during Neonatal ECMO, Transfusion, DOI: 10.1111/trf.14151.
• Ghosh, Santu, Chatterjee, Arpita, and Ghosh, Samiran (2017): Non-inferiority test based on transformations for non-normal distributions, Computational Statistics and Data Analysis, 113, 73-87.
• Ghosh S and Polansky A (2016): New Bootstrap Confidence Intervals for Means of Positively Skewed Distributions. Communication in Statistics-Theory and Methods, 45, 6915-6927.
• Polansky A and  Ghosh S (2015): Using Observed Confidence Level to Perform Principal Component Analyses. Communication in Statistics-Theory and Methods, 45, 3596-3611.
• Ghosh, Samiran, Ghosh Santu, and Tiwari, Ram(2015) : Bayesian Approach for Assessing Non-inferiority in a Three-Arm  Trial With Pre-Specified Margin. Statistics In Medicine, 35: 695–708.
• Ghosh S and Polansky A (2014): Smoothed and Iterated Bootstrap Confidence Regions for Parameter vectors. Journal of Multivariate Statistics, 142:171-182.