top of page
1894940_7561.jpg

Publications

Savita Pareeks Publications

1. Likelihood-based missing data analysis in crossover trials

Pareek, S.,  Das K., and Mukhopadhyay S., Brazilian Journal of Probability and Statistics, 37(2), Jun 2023. GitHub codes

Pareek, S., Khetan, M., Mukhopadhyay, S. and Das, K., JSR (Special Issue) Vol 55 Number 1 (2021).

3. Likelihood-based Inference for Skewed Responses in a Crossover Trial Setup

4. Freight Costs of India’s Trade

Biswas, D., Pareek S., and Saggar S., RBI Bulletin June 2022.

5. Capturing e-commerce transactions from the travel group of BPM6: an Indian experience

2. Clustering gene expression time series data embedded in a nonparametric setup

Pareek, S.,  Das K., and Mukhopadhyay S. (Under Review)  GitHub Codes, arXiv link

Misra, A, Pareek, S., and Saggar, S., External Statistics conference organised by the Irving Fisher Committee with the European Central Bank, in collaboration with the Bank of Spain, Madrid, 2024.

Post-doc Research Projects (Ongoing)

Savita Pareek Research

Ph.D. Research Projects

Mixed Effects Models (MEM) for Crossover Data
• Devised methods for analyzing crossover trials with incomplete or skewed data distributions using MEM.
• Applied Monte Carlo Expectation Maximization (MCEM) within MEM for maximum likelihood estimation of parameters with missing or skewed data.
•  Illustrated statistical modelling and estimation using gene expression data from a 3 ×3 crossover study.

Clustering Multiple Time-series Data
•  Developed model-based gene expression time series clustering using MEM and non-parametric Dirichlet processes to capture temporal effects.
•  Analyzed gene expression data from antibiotic-producing Streptomyces coelicolor using these techniques.


 Wavelet-based estimation in mixed-effect models

  • Working on developing a robust generalised method of wavelet moments for large longitudinal data.

  • ​Exploring model selection properties under mixed effect model setup.


 Dynamic Copula model estimation 

  • Devising dynamic copula based optimal portfolio schemes for financial returns.

 

 Differential privacy for compositional data 

bottom of page