Contact Information
Research Interests
Bayesian Neural Networks and Model Compression, Transfer Learning and Synthetic Data Augmentation, Bayesian Sparse Additive Modeling, Statistical Learning, Predictive Modeling and Anomaly Detection
Research Description
My research focuses on advancing the fields of Bayesian neural networks, transfer learning, and predictive modeling to address complex, high-dimensional data challenges in real-world applications.
Currently, I am collaborating with Dr. Feng Liang to develop Bayesian neural network models that incorporate novel model compression and feature selection techniques to enhance interpretability and computational efficiency. Our work specifically targets applications in mortality prediction and agricultural forecasting. Additionally, I am working with Dr. Yuexi Wang to explore innovative source-free semi-supervised Bayesian transfer learning approaches, aiming to improve predictive accuracy for small-target datasets by transferring knowledge from larger, unlabeled sources. Furthermore, my collaborator and I have developed an efficient variational Bayesian algorithm for simultaneous smoothing and feature selection in additive models.
My research also encompasses causal inference methodologies, where I have applied meta-learning to quantify causal effects across demographic subgroups, thereby enhancing precision in healthcare trial outcomes and resource allocation.
Prior to pursuing my Ph.D., I worked in Statistical Analysis at WalmartLabs, where I contributed to the development of an automated assortment machine for Walmart store owners in the UK and Canada.
Education
Ph.D. Candidate in Statistics
University of Illinois at Urbana-Champaign
Master of Statistics (M.Stat)
Indian Statistical Institute, Kolkata, 2017
Bachelor of Statistics (B.Stat)
Indian Statistical Institute, Kolkata, 2015
Awards and Honors
Summer Research Fellow, Predictive Analytics & Risk Management, UIUC, 2024
Recipient of the prestigious summer PARM fellowship for my research proposal on the use of Bayesian Neural Network methodologies in Human Mortality Data.
NSF Travel Grant for Student Oral Paper Competition, IISA, 2022
Recipient of the NSF travel grant in recognition of my presentation at the student oral paper competition at the International Indian Statistical Association (IISA) conference.
Outstanding Qualifying Exam Award, University of Illinois, Urbana Champaign, 2020
Recipient of the Outstanding Qualifying Exam Award for being the top student in my Ph.D. cohort at the University of Illinois, Urbana Champaign.
Excellent Teaching Assistant Recognition, University of Illinois, Urbana Champaign, 2019
Acknowledged as an Excellent Teaching Assistant for my contributions to enhancing the learning experience of students in various courses at the University of Illinois, Urbana Champaign.
Best Paper Award, National Conference on Machine Learning and Artificial Intelligence (NCMLAI), 2018
Awarded Best Paper for my research presented at the National Conference on Machine Learning and Artificial Intelligence.
Patent Filings, Walmart Labs, 2018/2019
Filed three patents related to advancements in statistical methodologies during my tenure at Walmart Labs.
Associate of the Month, Walmart, July 2018
Honored as Associate of the Month among over 2,000 associates for outstanding performance and contributions at Walmart.
INSPIRE, 2012 - 2017
Recipient of the Inspire Scholarship (given by the Department of Science and Technology, Govt of India).
Courses Taught
Instructor:
- 2021 Summer: Qualifying Exam Preparation Course
Discussion Sections (Lab Sessions):
- 2019 Fall: STAT 400 - Statistics and Probability 1
- 2024 Summer: STAT 400 - Statistics and Probability 1
Teaching Assistant:
- 2024 Fall: STAT 542 - Practical Statistical Learning
- 2023 Fall: STAT 542 - Practical Statistical Learning
- 2024 Spring: STAT 528 - Advanced Regression Analysis 2
Additional Campus Affiliations
-
Elected Member, Teaching Assistant Welfare Committee
Statistics Department, University of Illinois at Urbana-Champaign, 2022 -
Chair, Seminar Committee
Statistics Doctoral Student Association, University of Illinois at Urbana-Champaign, 2022
External Links
Highlighted Publications
-
Efficient Model Compression for Bayesian Neural Networks [Link]
Diptarka Saha, Zihe Liu, Feng Liang; 2023+;(Under Submission) -
Bayesian Smoothing and Feature Selection Using Variational Automatic Relevance Determination [Link]
Zihe Liu, Diptarka Saha, Feng Liang; 2023+; (Under Submission) -
Machine Learning-Assisted Array-Based Detection of Proteins in Serum Using Functionalized MoS2 Nanosheets and Green Fluorescent Protein Conjugates[Link]
Pradipta Behera, Krishna Kumar Singh, Subhendu Pandit, Diptarka Saha; ACS Applied Nano Materials 2021, 4 (4), 3843-3851; DOI: 10.1021/acsanm.1c00244. -
A Stepwise Approach to Automate the Search for Optimal Parameters in Seasonal ARIMA Models [Link]
Manisha Mukherjee, Diptarka Saha; Proceedings of Alliance International Conference on Artificial Intelligence and Machine Learning; 26 - 27 April 2019; ISBN: 978-93-5361-299-3. -
A Statistical Approach to Quantify Promotion Corrected Measure of Item Loyalty [Link]
Diptarka Saha; Proceedings of Alliance International Conference on Artificial Intelligence and Machine Learning; 26 - 27 April 2019; ISBN: 978-93-5361-299-3. -
REDCLAN: RElative Density based CLustering and ANomaly Detection [Link]
Diptarka Saha, Debanjana Banerjee, Bodhisattwa P. Majumder; 4th International Conference on Software Engineering; September 2018; Copenhagen, Denmark; ISBN: 978-1-921987-91-5; DOI: 10.5121/csit.2018.81303.