Sean M. Tatman

Analyst

Mr. Tatman has experience in modeling and simulations. He is the project manager for the IDI team simulating shipping ports using ExtendSim to analyze the impact of interruption to operations due to natural disasters, etc.

He was part of a machine learning project to predict future water pipe failures. The work involved discretization techniques and Bayesian Network structure learning tools in Python and R. As part of this project he also performed analysis and visuals for model calibration metrics.

He assisted in the model simulation of a distributed fueling network for the Navy in ExtendSim.

He was a team member on 1 out of 25 teams selected for the AI Health Outcomes Challenge competition created by the Center for Medicare and Medicaid. The challenge involved a large amount of data hosted on a PostgreSQL database. Feature engineering and models were created using SQL and Python.

Mr. Tatman has evaluated the predictive performance of inference enterprise models using certainty intervals by recreating the models and running it on simulated population scenarios. Inference enterprise models use a variety of detectors to predict typical employee behaviors related to insider threats. Tools for the enterprise models have included Keras a deep learning python library and DBSCAN a clustering package/algorithm in R.

He also analyzed Anomaly Detection in Bio-Surveillance using Bayesian Networks in Netica and R. He Supported the Route Reconnaissance and Clearance Training Simulator Business Case Analysis (R2C Sim BCA), aiding in the development of the mathematical optimization tool that generated the best portfolio of training technologies based on a myriad of training requirements by different engineering units at varying Marine Corps locations.

Mr. Tatman also has experience developing instructional materials and documenting code.  He prepared instructional material (Powerpoint slides and class examples) for machine learning and discretization of continuous variables for “Bayesian Networks using Netica” professional short course. He created coaching materials to aid a customer in creating diagnostic models, e.g. diesel generators. He created a quickly updateable developer’s guide for a Real Time Evacuation Planning Model (RtePM) with the Virginia Modeling, Analysis, and Simulation Center (VMASC), and developed documentation for the verification and validation phase of the project.

Mr. Tatman has an M.S. in Data Analytics Engineering at George Mason University. As part of his coursework, he has analyzed the benefits of NetGalley, a service for the publishing industry to gain reviews of books before release. He performed statistical analysis to model fast-food chain restaurant’s revenue based on location and other factors in SAS programming language. He also participated in a design exercise building an analytical tool to combat gang violence.

Sean is a Certified Analytics Professional (CAP) with the INFORMS professional association.

Education

M.S Data Analytics Engineering, George Mason University, (2017)
B.S. Mechanical Engineering, University of Virginia, 2011

Selected Publications and Presentations

Russell Mosier, Ronald Woodaman, Sean Tatman, and Jonathan York. “Decision Analysis and Optimization Methodologies Employed in the 2012 Marine Corps Route Reconnaissance and Clearance Training Simulators Business Case Analysis.” 81st Military Operations Research Society Symposium, Crystal City, VA, June 2013.

Tatman, J. Hepler, A. Smith, G. Tatman, S. et al. “Combining Technology and Soft Skills in the Bayesia Expert Knowledge Elicitation Environment (BEKEE)”, INFORMS 2014, TD 61, pp 350.