Data Science, Economics
Resume:
DATA SCIENTIST
TRAVIS
JAMES
á ************@*****.***
â travis-james
Education
The London School of Economics
and Political Science
MSc Economics 2016
The University of Washington
B.S. (Honors, Cum Laude)
Economics 2014
Dean's List Every Academic Quarter
Minors in Mathematics and Spanish
Skills
CODING LANGUAGES
Python
SQL
Stata
STATISTICS
Hypothesis Testing
Bayesian Inference
Regression Analyis
Time Series
Probability Theory
Panel Data
MACHINE LEARNING
Cross Validation
Grid Search
Clustering
PCA
Neural Networks
Natural Language Processing
SPOKEN LANGUAGES
English (native speaker)
Spanish (conversational uency)
Activities
Rationale Magazine
Editor and Staff Writer - LSE
The Economics Undergraduate Board
The University of Washington
Teaching Assistant
Greg Ellis - The University of Washington
The Economizer
Writer - The University of Washington
Global Right to Education
Applicant Mentor - London, United Kingdom
G.R.U.B. Volunteer - Olympia, WA
Experience
Metis San Francisco, CA
Data Science Fellow Sep 2016 to Dec 2016
Part of a 12 week immersive data science bootcamp run through Kaplan, Inc. Developed 5 end-to-end data science projects utilizing exploratory data analysis, supervised and unsupervised machine learning, predictive analytics, natural language processing, data collection and munging, and database selection and management.
Pitchbook Data, Inc. Seattle, WA
Research Analyst Apr 2014 to Jan 2015
Conducted private equity and venture capital research and database management tasks for the research division.
Wrote SQL queries to pull relevant data from Pitchbook's proprietary database. Managed customer queries and developed strategies to solve them, while also managing team of interns. The Washington Business Alliance Seattle, WA
Research and Project Management Intern Jun 2013 to Aug 2013 Conducted economic research under Lisa Wellman, former VP of Apple’s publishing and new media division. Created a metric database of key public policy statistics within the state of Washington. Researched Washington state's standing within the elds of economic development, healthcare, education, environment and quality of life.
Progeny 3 Seattle, WA
Charted Financial Analyst Assistant Feb 2013 to Apr 2013 Worked directly with head of private equity investments on economic outlook of Indonesia. Utilized Bloomberg software and other sources to generate data and graphs. Presented actionable insights to nancial investment team. Enterpricing Buenos Aires, Argentina
International Tax Analyst Assistant Jun 2012 to Aug 2012 Conducted transfer pricing studies and completed economic outlooks on industries within Argentina. Performed contractual research between companies.
Utilized LexusNexis database for relevant data and information. Projects
Tracking Trends in Innovation Current
Used NLP to analyze weekly patent grant lings and visualize trends among innovation topics over time. Also used PageRank algorithm to show importance of patents based on citations, and turned model into interactive web app using Flask.
Predicting Loan Application Decision and Grade
Used Lending Club data to predict loan application decision and grade. Also built a ask web application to make model interactive, and included a d3.js visualization of the United States to demonstrate state-level loan statistics. Application decision model has an accuracy of 94%, precision of 82% and receall of 80%. Application grade model has an accuracy of 56% (random guessing yields 20%). Predicting IMDB Movie Review Sentiment
Predicted IMDB user scores for various movies using supervised machine learning techniques. These included both parametric regressions and non-parametric methods such as random forests and gradient boosting. Gradient boosting model has a test set adjusted R-squared of 0.6879. Run Time and Metascore positively in uence public sentiment. Oscar Wins and Budget negatively in uence public sentiment. Street Performer Allocation
Used New York City Metro data to perform an exploratory data analysis and came up with a sorting algorithm to equitably distribute street performers throughout the various subway stations of NYC. A Formal Comparison of Religious Texts
Used NLP topic modeling to compare the key themes of ve major religious and philosophical texts. Also performed sentiment analysis to compare the subjectivity and polarity of texts.
Contact this candidate