Current Research Interests

  • Statistics on non-Euclidean object-valued data- developing broadly applicable statistical methods and inference for analyzing non-Euclidean data residing in abstract metric spaces, with applications in brain imaging studies, mortality and life expectancy distributions, child neurological development, traffic network analysis, and genetics data.
  • Functional and longitudinal data analysis, and its overlap with metric geometry for studying samples of time-varying metric space-valued data, examples being dynamic networks or distribution objects
  • Reproducing kernel Hilbert spaces- developing abstract mathematical and computational methods for infinite dimensional data in connection with metric geometry and sufficient dimension reduction.
  • Nonparametric statistics, analysis of high-dimensional and geometrical data.
  • Causal inference for distributional and object data- developing methodological work for object data analysis with application in the fields of bio-sciences.
  • Theoretical machine learning- contextual bandit problems.