I research in a wide variety of disciplines, applying quantitative, computing and mathematical skills in various areas. My research interests are inter-disciplinary in nature, spanning the areas of:
- Statistics and quantitative methods • Marketing Research
- Management Science and Operations Research
- Machine learning, Artificial Intelligence and programming principles (especially C++ and Python with their applications to sustainable AI)
- Social Research, including research in the fields of criminology, psychology and sociology
- In addition to these, my work has also been published in the areas of health.
My PhD focussed on theoretical development and programming of statistical and machine models for heterogeneity estimation using variational Bayes, which is an Artificial Intelligence approach for estimating Bayesian models without Markov Chains Monte Carlo methods. I proposed new models, in particular, Variational Bayesian approach to estimating polychotomous mixture models and also tested this technique on mixture regression models. I also applied theoretical models to the field of corporate innovation and introduced the concept of ‘innovation ambidexterity’ for enterprises.
My recent work modelled how engagement with various media sources shapes our perception about our environment and how this perception varies with proximity to this environment. I uncovered the role of the number of sources we engage with as being a key to this relationship by designing a statistical model around it. As part of this research, I proposed a novel concept of ‘repertoires of information’.
Currently, I am researching the viability of enterprise AI models in terms of their computational costs and the trade-offs that companies might have to make in the future while using these models. In this regard, I am looking at programming languages, such as C++, Python and Rust.