John Honaker Email and Phone Number

Perform statistical consulting, model validation, and independent review of statistical and machine learning models from all aspects of Nationwide. Primarily responsible for assessing and challenging the statistical methodology of modelling teams, communicating the risks of modelling approaches to governing risk committees, and assisting with the mitigation of any risks or issues discovered.

Develop a wide variety of statistical and machine learning models for the Enterprise Analytics Office. Some highlights include:- A natural language processing (NLP) model identifying subrogation opportunities in incoming claims. This lead to an increase in identified recovery of $16M to $80M- A per-trip-level risk association model that distinguished the risk of an auto accident based on the actual route driven.- Co-authored a whitepaper establishing NLP best practices for the department.- Created a method for setting quantitative, data-based goals and expectations for customer enthusiasm surveys.- Developed a Shiny dashboard for ongoing process monitoring compatible with Teradata SQL databases.

Data analysis, study design, scientific writing, explaining statistical methods to other researchers

Statistical analysis for various research projects. SAS programming. R programming. Compiling literature reviews.