Main
Katherine Goode
Statistician
My research focuses on the explainability and interpretability of machine learning models. Other research interests include model assessment, data visualization, and random forest models. My full CV is available here.
Education
Iowa State University
Ph.D in Statistics
Ames, Iowa
2021
Thesis: Visual diagnostics for explaining machine learning models
University of Wisconsin, Madison
M.S. in Statistics
Madison, Wisconsin
2015
Lawrence University
B.A. in mathematics
Appleton, Wisconsin
2013
Experience
Senior Member of Technical Staff
Statistical Sciences, Sandia National Laboratories
Albuquerque, NM
Current - 2021
Research and development of statistical methods in application areas including climate and cyber security
Postdoctoral Researcher
Statistical Sciences, Sandia National Laboratories
Albuquerque, NM
2021
Researched use of elastic shape analysis with inverse models for functional data using and developed feature importance technique for echo state networks applied to climate data
Research and Development Intern
Statistical Sciences, Sandia National Laboratories
Albuquerque, NM
2021 - 2019
Developed explainable machine learning pipeline for functional data Presented on work at internal and external events
Graduate Research Assisstant
Natural Resource Ecology and Management, ISU
Ames, Iowa
2021 and 2019
Developed R Shiny application to predict taxonomy of fish eggs using random forests (2021) and assisted with analysis of toxicology study of monarch butterfly larvae exposed to insecticides (2019)
Statistical Consultant
Agriculture Experiment Station, ISU
Ames, Iowa
2020 - 2016
Provided statistical support on research projects for graduate students, professors, and staff across university departments
Software
listenr
R package
2024
Echo state networks with spatio-temporal feature importance
redres
R package
github.com/goodekat/redres.git
2019
Residuals and diagnostic plots for mixed models.
ggResidpanel
R package
CRAN.R-project.org/package=ggResidpanel
2019
Panels and interactive versions of diagnostic plots using ggplot2.