Joshua Proctor

Sr. Research Scientist

Joshua Proctor

Sr. Research Scientist


Professional Appointments

Affiliate Assistant Professor, Applied Mathematics, University of Washington Affiliate Assistant Professor, Mechanical Engineering, University of Washington

Joshua Proctor is currently a research manager and senior research scientist at IDM leading the Data, Dynamics and Analytics (DDA) team. DDA is focused on applying modern data science, machine-learning, and statistical techniques to a wide-range of data sources, i.e., surveillance of infectious diseases, household surveys, genome sequencing of viruses or parasites, and demography.  We aim to leverage the growing success of these modern analytic and numerical methods to advise on near-term public policy questions facing the global health community. In conjunction, we also identify the short-comings of current methodological approaches and develop novel, principled algorithms to face the challenges of surveillance data.

Joshua Proctor earned a Ph.D. in Mechanical and Aerospace Engineering from Princeton.  Before graduate school, Joshua earned a Bachelor of Science in Aeronautics and Astronautics Engineering, and a Bachelor of Arts in English Literature, both from the University of Washington, Seattle. His doctoral research focused on investigating the effects of neural feedback on rapidly running insects (specifically, cockroaches).  The research required the development of complex mathematical models describing legged locomotion, the application of dimensionality reduction techniques, and the characterization of these nonlinear dynamical systems through bifurcation analyses [1,2].  The research led to several important discoveries about the role of neural feedback during running, while also inspiring better robotic designs for maneuverability, stability, and control.

Joshua joined IDM in 2011 after his Ph.D. and contributed to the development of the compartmental modeling simulations software.  This software package is soon to be released to the global health community.  Joshua then joined the nascent Applied Mathematics group at IDM where he developed algorithms and novel mathematical methods for the study of infectious disease data.  He became interested in equation-free modeling and data-driven analyses, inspired by the mathematical developments around fluid dynamics at Princeton University.  Generally, equation-free modeling does not require a set of pre-determined, derived from first-principles equations.  Instead, the time-series data is utilized to discover dynamic models and/or dynamic characteristics.  This led to a number of important methodological innovations [3,4,5,6], including a book describing the mathematical method Dynamic Mode Decomposition [7]. These methodologies are poised to have a substantial impact on infectious disease modeling.

During his previous position in the Applied Mathematics group, Joshua also focused on leveraging current data science and machine-learning methodologies for epidemiological and demographic applications.  For example, he is interested in characterizing genomes of viruses or parasites and their metadata (GPS location, age, gender, sex, etc.) to better understand the transmission dynamics of infectious diseases [8,9]. Understanding these modern genomic data sets can help inform disease surveillance efforts for elimination and eradication efforts.  Joshua is also interested in questions around demography, specifically in the family planning group and estimating child mortality.

Biography

Professional Appointments

Affiliate Assistant Professor, Applied Mathematics, University of Washington Affiliate Assistant Professor, Mechanical Engineering, University of Washington

Joshua Proctor is currently a research manager and senior research scientist at IDM leading the Data, Dynamics and Analytics (DDA) team. DDA is focused on applying modern data science, machine-learning, and statistical techniques to a wide-range of data sources, i.e., surveillance of infectious diseases, household surveys, genome sequencing of viruses or parasites, and demography.  We aim to leverage the growing success of these modern analytic and numerical methods to advise on near-term public policy questions facing the global health community. In conjunction, we also identify the short-comings of current methodological approaches and develop novel, principled algorithms to face the challenges of surveillance data.

Joshua Proctor earned a Ph.D. in Mechanical and Aerospace Engineering from Princeton.  Before graduate school, Joshua earned a Bachelor of Science in Aeronautics and Astronautics Engineering, and a Bachelor of Arts in English Literature, both from the University of Washington, Seattle. His doctoral research focused on investigating the effects of neural feedback on rapidly running insects (specifically, cockroaches).  The research required the development of complex mathematical models describing legged locomotion, the application of dimensionality reduction techniques, and the characterization of these nonlinear dynamical systems through bifurcation analyses [1,2].  The research led to several important discoveries about the role of neural feedback during running, while also inspiring better robotic designs for maneuverability, stability, and control.

Joshua joined IDM in 2011 after his Ph.D. and contributed to the development of the compartmental modeling simulations software.  This software package is soon to be released to the global health community.  Joshua then joined the nascent Applied Mathematics group at IDM where he developed algorithms and novel mathematical methods for the study of infectious disease data.  He became interested in equation-free modeling and data-driven analyses, inspired by the mathematical developments around fluid dynamics at Princeton University.  Generally, equation-free modeling does not require a set of pre-determined, derived from first-principles equations.  Instead, the time-series data is utilized to discover dynamic models and/or dynamic characteristics.  This led to a number of important methodological innovations [3,4,5,6], including a book describing the mathematical method Dynamic Mode Decomposition [7]. These methodologies are poised to have a substantial impact on infectious disease modeling.

During his previous position in the Applied Mathematics group, Joshua also focused on leveraging current data science and machine-learning methodologies for epidemiological and demographic applications.  For example, he is interested in characterizing genomes of viruses or parasites and their metadata (GPS location, age, gender, sex, etc.) to better understand the transmission dynamics of infectious diseases [8,9]. Understanding these modern genomic data sets can help inform disease surveillance efforts for elimination and eradication efforts.  Joshua is also interested in questions around demography, specifically in the family planning group and estimating child mortality.

Publications

Saturday, October 21, 2017
Dynamic mode decomposition has emerged as a leading technique to identify spatiotemporal coherent structures from high-dimensional data, benefiting from a strong connection to nonlinear dynamical syste
Read online
Tuesday, May 30, 2017

​We present a universal, data-driven decomposition of chaos as an intermittently forced linear system. ​​

Read online
Wednesday, April 26, 2017
The method provides a promising new technique for discovering governing equations and physical laws in parameterized spatiotemporal systems. ​
Read online
Tuesday, April 25, 2017

Containing the recent West African outbreak of Ebola virus (EBOV) required the deployment of substantial global resources.

Read online
Thursday, January 19, 2017

​We propose an alternative data-driven method to infer networked nonlinear dynamical systems by using sparsity-promoting optimization to select a subset of nonlinear interactions representing dynam

Read online
Tuesday, January 17, 2017
The integration of nonlinear dynamics and machine learning opens the door for principled versus heuristic methods for model construction, nonlinear control strategies, and sensor placement techniques.
Read online
Wednesday, January 11, 2017
Using a computational model of the Caenorhabditis elegans connectome dynamics, we show that proprioceptive feedback is necessary for sustained dynamic responses to external input. ​
Read online
Friday, January 6, 2017
We develop an algorithm for model selection which allows for the consideration of a combinatorially large number of candidate models governing a dynamical system.
Read online
Thursday, December 1, 2016
​This work develops compressed sensing strategies for computing the dynamic mode decomposition (DMD) from heavily subsampled or compressed data. ​
Read online
Tuesday, November 22, 2016

​We review standard model reduction techniques such as Proper Orthogonal Decomposition (POD) with Galerkin projection and Balanced POD (BPOD). ​

Read online
Wednesday, October 26, 2016

​Using recent advances in sparsity-promoting techniques, we present a novel algorithm to solve this sparse sensor placement optimization for classification (SSPOC) that exploits low-dimensional str

Read online
Sunday, July 24, 2016
We consider the application of Koopman theory to nonlinear partial differential equations
Read online
Wednesday, February 24, 2016

We develop a new generalization of Koopman operator theory that incorporates the effects of inputs and control.

Read online
Tuesday, January 26, 2016
This publication explores a new method, called Dynamic Mode Decomposition with control (DMDc), which extends DMD to incorporate the effect of inputs and control.
Read online
Saturday, January 16, 2016
We explore finite-dimensional linear representations of nonlinear dynamical systems by restricting the Koopman operator to an invariant subspace spanned by specially chosen observable functions. ​
Read online
Sunday, October 11, 2015
Koopman spectral analysis provides an operator-theoretic perspective to dynamical systems rather than the more standard geometric and probabilistic perspectives.
Read online
Friday, September 11, 2015
Extracting physical laws from data is a central challenge in many diverse areas of science and engineering.
Read online
Thursday, May 7, 2015

Traditional methods for estimating malaria transmission based on mosquito sampling are not standardized and are unavailable in many countries in sub-Saharan Africa.

Read online
Sunday, March 1, 2015
The development and application of quantitative methods to understand disease dynamics and plan interventions is becoming increasingly important in the push toward eradication of human infectious dise
Read online
Wednesday, December 10, 2014
Complex systems exhibit dynamics that typically evolve on low-dimensional attractors and may have sparse representation in some optimal basis.
Read online
Thursday, October 2, 2014

The eradication of diseases has long been a focus of the global health community and researchers in epidemiology and other related fields.

Read online
Monday, September 22, 2014
We develop a new method which extends Dynamic Mode Decomposition (DMD) to incorporate the effect of control to extract low-order models from high-dimensional, complex systems.
Read online
Thursday, April 17, 2014
The haemozoin crystal continues to be investigated extensively for its potential as a biomarker for malaria diagnostics.
Read online
Wednesday, December 18, 2013

This work develops compressive sampling strategies for computing the dynamic mode decomposition (DMD) from heavily subs

Read online
Tuesday, October 15, 2013
The goal of compressive sensing is efficient reconstruction of data from few measurements, sometimes leading to a categorical decision.
Read online