SAURABH BHARDWAJ
Thapar Institute of Engineering and Technology (Deemed to be University), Virginia Tech
0000-0002-6718-9691
About
With over 20 years of academic and research experience, I currently serve as a Professor and Associate Dean of Faculty Affairs at the Thapar Institute of Engineering and Technology in India. Additionally, I am a Research Scientist at Virginia Tech, USA, where I collaborate on pioneering bioinformatics projects.
My research primarily focuses on advancing statistical and machine learning methodologies, with a specialization in bioinformatics. In my recent work, I contributed to the development of an analytic tool suite designed to identify informative molecular features across different phenotypic groups. This suite supports essential tasks such as missing value imputation, signature gene detection, and expression pattern visualization, specifically tailored for analyzing biologically diverse samples. Key innovations include mechanism-integrated group-wise imputation for identifying signature genes with informative missingness, an extended cosine-based one-sample test to detect enumerated signature genes, and a unified heatmap for comparative display of complex expression patterns.
I also contributed to the development of the fully unsupervised dCAM, an iterative algorithm that simultaneously estimates cell type-specific gene expression profiles and cell type proportions while performing differential expression analysis at the cell type level.
In addition, I am part of a team working on the blind separation of dependent, non-negative sources in various applications. The CAM framework introduces significant advancements in blind source separation, particularly through simplex characterization, and offers the first unsupervised deconvolution method for complex datasets.
Furthermore, I have explored the Cosine-based One-Sample Test (COT) in scatter space to detect marker genes across subtypes using subtype expression profiles, the Sample Progression Discovery method for identifying biological progression patterns in microarray gene expression data, and the Differential Dependency Network (DDN) tool for detecting and visualizing statistically significant topological changes in transcriptional networks across different biological conditions.
My previous research includes applying machine learning and deep learning techniques in fields such as speaker identification, brain fingerprinting, clustering, and solar radiation prediction. My academic background includes a Postdoctoral Fellowship at Virginia Tech and advanced degrees in instrumentation engineering from Panjab University and the Netaji Subhas Institute of Technology, Delhi University.
Employment
-
Thapar Institute of Engineering and Technology (Deemed to be University) Professor and Associate Dean Faculty Affairs2014 - Present
Education
-
Virginia Tech PostDoc2020 - 2022
Projects & Funding
Projects & funding information is unavailable.
Publications (32)
- An efficient speaker identification framework based on Mask R-CNN classifier parameter optimized using hosted cuckoo optimization (HCO) Save
- An Intelligent Approach of Measurement and Uncertainty Estimation for Hidden Information Detection Using Brain Signals Save
- A Review for the Development of ANN Based Solar Radiation Estimation Models Save
- Accuracy Improvement of Solar Power Estimation Using Real-Time Degradation Computation of PV Panels Save
- An Artificial Neural Network Based Approach of Solar Radiation Estimation Using Location and Meteorological Details Save
- Global Solar Radiation Estimation Modeling Using Artificial Neural Network: A Case Study on Metro Cities of India Save
- Information Detection in Brain Using Wavelet Features and K-Nearest Neighbor Save
- An information set-based robust text-independent speaker authentication Save
- Artificial Neural Networks Based Solar Radiation Estimation using Backpropagation Algorithm Save
- Classification of EEG signals using hybrid combination of features for lie detection Save