1. Multivariate signal processing
2. Diagnostic predictive modeling
3. Data-centric machine learning
I did my undergraduate work at the University of Arizona and obtained my Master's degree from Stanford University, both in Electrical Engineering. I then worked at Mosys Inc., California, where I designed analog circuits for SERDES. Upon my return to India I worked as a Research Associate at IIT Delhi and went on to obtain my doctorate degree from there. My work at IIT Delhi was on engineering novel computational biomarkers from neural signals using signal processing and statistical learning. After completing my PhD I cofounded a company, Intellihealth, with the aim of providing the technologies I work on in the hands of patients and clinicians as quickly as possible.
I work at the intersection of signal processing, machine learning and diagnostic medicine. My approach can be termed "data-centric machine learning" where the attempt is to trade-off model complexity and data size with data engineering through signal processing. In medical applications, obtaining large amounts of high quality data is often not feasible. By understanding the clinical context and engineering the data to extract the information where the biomarkers of the disease are residing one can solve many challenging diagnostic problems. I work mostly on, but not restricted to, neurological disorders.