Worked with a team of research scientists to develop models in Python to understand medical datasets on patients infected with COVID-19. The data consists of 75 features for which XGBoost was used to select those features which are most important. We also created a function in Pytorch that takes data from blood samples of patients as input and outputs a survival probability with a validation accuracy of 93.9%.

Assisted in grading projects and exams and providing detailed feedback to students taking a postgraduate course in statistics. Projects required the use of technical software such as R and Python.

Managed and lectured courses to 40+ undergraduate students including multivariable calculus, differential equations, probability & statistics, statistical programming in Matlab, and Mathematica.

Conducted research on the econometrics of network formation. In most of the real world social and economic networks, there are some common features, which are widely observed. “Homophily”, which refers to the tendency of individuals to make connections or friendships with other like-minded individuals, and “degree heterogeneity”, which refers to the variation in the total number of links per individual, are some of the common features. The presence, absence, and magnitude of such features influence information diffusion, the spread of epidemics, and social learning procedures on the underlying networks.