I'm interested in data-driven manifold learning and dimensionality reduction methods specifically in the context of multiscale, stochastic, high-dimensional complex systems. My work focuses on developing algorithmic and theoretical aspects of optimal design of measurements and inputs for forecasting and control. This optimization relies on model reduction to obtain a low-rank representation of dynamics. I have applied this framework to sensor placements in fluid flows, ocean temperature data, aircraft shimming data, and more generally to optimal sensor and actuator placement for linear time invariant systems.



Data-driven learning 

Application of machine learning to real-world, high-dimensional data with the goal of replicating underlying dynamical phenomenon for prediction:

Dimensionality reduction, manifold learning, operator theoretic approaches (i.e. Koopman)

2013 - 2018

University of Washington

PhD Applied Mathematics

Optimal design

Structuring measurements and inputs for optimizing learning, inference and control objectives:  sensor/actuator placement, sensor fusion, sampling


Aerodynamics and flows

Smart manufacturing and data processing

Geophysics and climate

2009 - 2013

University of Massachusetts Lowell

BSc Mathematics and Computer Science

2007 - 2009

Massachusetts College of Art & Design


  • Facebook Clean Grey
  • Twitter Clean Grey
  • LinkedIn Clean Grey

© 2018 by Krithika Manohar. Proudly created with