The Walsh Data Science Laboratory
Stephen J. Walsh, Ph.D.
Data Science, Machine Learning, Applied and Mathematical Statistics
Born: December 10, 1981 (40yo)
Citizenship: United States of America
All images ©COPYRIGHT
by
Stephen J. Walsh
2020
All Rights Reserved
Employment:
Assistant professor of statistics at Utah State University, College of Science, Department of Mathematics and Statistics
Dissertation:
Development and Applications of Particle Swarm Optimization for Constructing Optimal Experimental Designs.
https://scholarworks.montana.edu/xmlui/handle/1/16309
Contact Info:
walsh.datascience "at" gmail.com
steve.walsh "at" usu.edu
Summary
Skilled data scientist, statistician and machine learner with a broad applied experience and deep mathematical expertise. Demonstrated interdisciplinary practice in fields:
Materials Science
Physics
Chemistry
Metrology, method development and validation, GUM uncertainty methodology
Laboratory quality control and quality assurance according to international standards (e.g. ISO)
Managing, analyzing, and reporting the results of interlaboratory comparisons and proficiency tests
Remote Sensing and image analysis
Nuclear engineering, international safeguards, and various applications
Versatile data-science programming capabilities:
High level proficiency in languages: Julia, R (base and tidyverse), MATLAB, Mathematica, Sage, LaTex, R-markdown
Basic proficiency in: SAS, Minitab, Python
Proficiency in all Microsoft office and google office products
Deep knowledge and proficiency with the following statistics and machine learning sub-disciplines:
Response Surface Methodology (product and process optimization via designed experiments)
Design of Experiments (DoE). Well versed in classical (catalogue) designs, optimal DoE in the hypercube, hypersphere, and standard n-simplex (i.e. mixture experiments).
Statistical quality and process control, including capability and tolerance analysis
Advanced mathematical-statistics training, advanced matrix algebra (e.g. vector spaces and projection operators), advanced matrix calculus (e.g. via Kronecker product), advanced inference theorems
Statistical modeling and inference including: linear and non-linear modeling and analysis, generalized linear models, mixed-effects models, and general likelihood based inferential procedures.
Statistical sample plan design
Robust statistical estimation methods, including estimation of Gaussian parameters, robust estimation of linear and mixed-effects model parameters, robust multivariate analysis methods, including characterization of multivariate Gaussian distributions, robust estimation of covariance matrices, and robust K-means type clustering algorithms. (outliers are a disaster to your data analysis, btw ;-) )
Multiple multivariate analysis techniques (e.g. PCA; clustering; factor analysis; MANOVA; discrimination and classification techniques such as LDA/QDA, SVMs, and mixture-distribution modeling)
Exploratory data analysis and graphical analysis
Spatial data analysis. General time-series analysis and state-space models
Bayesian modeling, sampling, inference, and data-analysis; Bayesian model averaging; Bayesian Networks
Trained in Biostatistics
Missing data analysis including Expectation-Maximization and Bayesian Data Augmentation
Multiple optimization techniques and algorithms including: gradient search with boosting, constrained optimization, Newton-Raphson, Fisher Scoring, Gauss-Newton, and of course meta-heuristics including the Particle Swarm Optimization