COLLIN WITT
Gaithersburg,MD
Summary
Mathematical Statistician with over a decade of advanced methodological research, model development, and statistical production.
Skilled statistician with strong track record in data analysis, statistical modeling, and problem-solving. Adept at collaborating with teams to achieve meaningful results and adapt to changing project needs. Proficient in statistical software, data visualization, and predictive analytics. Known for reliability, precision, and clear communication.
Overview
13
13
years of professional experience
Work History
Mathematical Statistician, Current Employment Statistics (CES)
Bureau of Labor Statistics
01.2020 - Current
- Produced National Variance Estimates for monthly production and special data requests supporting the CES Program Office.
- Evaluated and enhanced GEN3 and GEN4 forecasting frameworks using ETS, ARIMA, TRAMO REGARIMA, BaggedETS, and more.
- Improved Birth Death estimation by identifying overfitting in decision tree models and analyzing GLIMMIX, ROBUSTREG, and more.
- Developed and expanded the TS Breaks R Shiny app, integrating asynchronous processing, structural break detection, caching, and more.
- Provided internal customer support including responses to GAO inquiries and interpreting CES microdata structures, variance methods, and more.
- Collaborated with senior CES methodologists including Julie G., Greg Erkens, Chris Grieves, Steve Nance, and Dan Zhao on research projects.
- Mentored junior staff in Linux, SAS programming, microdata workflows, R modeling, and Shiny development.
Mathematical Statistician, Producer Price Index (PPI)
Bureau of Labor Statistics
01.2013 - 01.2019
- Led development of a Backup Sampling System in R Shiny and SAS, including frame clustering, sample selection interfaces, and more.
- Conducted Census frame research, resolved discrepancies in LBD and Census data, identified defects, and coordinated corrections.
- Developed SAS programs for employment and wage tabulation, regression analysis, weight smoothing, and many-to-many matching.
- Presented research at JSM 2016 on the use of Census data as a frame source for PPI; contributed to JSM Proceedings.
- Served as Lead Coder on the Backup Sampling System team, collaborating with analysts and methodologists to refine sampling methods.
- Participated in the Sample Reselection Team (BLS Innovation Award recipient), IPSONRRT, and COMPASS review group to improve processes.
- Mentored new employees in statistical analysis, sampling procedures, regression-based weighting, and system workflows.
Education
MS - Statistics
Florida State University
Tallahassee, FL05.2012
MS - Mathematics
College of Charleston
Charleston, SC05.2009
BS - Applied Mathematics
Coastal Carolina University
Conway05.2006
Skills
- R, SAS, Python, SQL, C, Linux, Shiny, Git
- Bayesian Methods, Time Series Forecasting, Isolation Forest, Random Forest, Gradient Boosting
- Survey Methodology, BRR and Variance Estimation, Regression and GLMMs, Sample Design
PUBLICATIONS
- Testing of Weight Smoothing Models in the Current Employment Statistics Survey with SAS and R (2023).
- Predicting the Effect of Business Births and Deaths on the Current Employment Statistics Survey (2023, with Chris Grieves and Steve Nance).
- The Efficacy of Using Economic Census Data as a Frame Source for Producer Price Index (2016).
CORE COMPETENCIES
- Customer Service; Decision Making; Mathematical Reasoning; Partnering; Team Building
- Bayesian Modeling; Machine Learning; Time Series Forecasting; R; SAS; Python; SQL; Shiny Development; Survey Methods; Variance Estimation
Timeline
Mathematical Statistician, Current Employment Statistics (CES)
Bureau of Labor Statistics
01.2020 - Current
Mathematical Statistician, Producer Price Index (PPI)
Bureau of Labor Statistics
01.2013 - 01.2019
MS - Statistics
Florida State University
MS - Mathematics
College of Charleston
BS - Applied Mathematics
Coastal Carolina University