Ryan Mcmahon Email & Phone Number

Denver, Co, Us

Denver, Co, Us

Boulder, Colorado, United States

Boulder, Colorado, United States

San Francisco Bay Area

YouTube Data Science team

Greater New York City Area

Applying natural language processing, machine learning, and other quantitative methodologies to assess the geo-political risks faced by businesses.

State College, Pennsylvania Area

This selective department level fellowship awards me a work release, which I am using to work on my dissertation and other co-authored projects.

2016/17: A research rotation concerning the application of Bayesian methods with large scale data -- i.e., when conventional MCMC estimation is computationally infeasible. With a focus on parallel processing approaches, we use latent variable models as a testing bed. Using simulation data, we conduct experiments on a High Performance Computing (HPC) cluster in order to determine the impacts of this method on model validity and computation time. We then provide an application of the methods to a large political survey. Initial findings suggest that these methods reduce model estimation time exponentially, while still accurately recovering the parameters.2015/16: A research rotation investigating links between media coverage of the United States economy and public opinion. We clean and analyze hundreds of thousands of articles from multiple news outlets across the last 60 years using a variety of natural language processing and supervised machine learning techniques. Additionally, we introduce Variance Weighted Classification (VWC) as a method for utilizing multiple codings of individual objects and show that this method yields higher model accuracy than using the modal coding or treating an object's codings as independent.

Programming in R and Python for Natural Language Processing using supervised machine learning methods, as well as conducting Optical Character Recognition of newspaper articles and data retrieval.

Teaching assistant for the main introductory course in political science.

Greater New York City Area

Using a corpus of over 1 billion tweets from survey respondents’ Twitter feeds, we study the link between exposure to political topics on Twitter during the 2016 election cycle and users’ opinions about those topics. Gained experience working in a map-reduce setting with PySpark and Hive.

Denver, Co

While at Slope I spent most of my time working on validating teacher and student metrics for the Colorado Department of Education. I developed a clear and functional workflow with regards to both my code and data management. Other projects included small projects for the analytics blog and exploratory data analysis for other startups. As a side project I collected and cleaned 37 years worth of Division 1 college basketball data. This was used to generate a measurement of the best college basketball teams in the shot clock era.

Coding statements of administration position (SAPs) and evaluating the SAPs' effects on the related legislation.

Doctor Of Philosophy (Ph.D.), Political Science (Methodology, American Politics; Grad Minor In Social Data Analytics)

My dissertation ("Say It With Style: The use and importance of linguistic style in political communications") examines how political.

Master Of Arts (M.A.), Political Science

My thesis was focused on the relationship between politics and finance. Specifically, I investigated the potentially causal effects.

Bachelor Of Arts (B.A.), Political Science, Cum Laude