Garrett Hoffman
| Data Scientist
Experience
Actuarial Consultant, Planning and Risk Management
Xerox Services
Secaucus, NJ
 · 
Jul 2011 to Current

  • Work closely with clients to identify issues and opportunities surrounding employee benefit programs and deliver financial/strategic recommendations.
  • Lead task force charged with brainstorming and prototyping  Data Analytics solutions for Human Resource Services. 
  • Perform valuations of liabilities in excess of $10 Billion through engineering of life contingency models  and identify actionable insights.
  • Forecast future funding and reporting requirements to develop budget recommendations and  risk management strategies .
  • Communicate quantitative results to non-technical supervisors and clients through presentations and written reports to inform decision making process.

Freelance Data Journalist
Priceonomics
San Francisco, CA (Remote)
 · 
Mar 2016 to Current

  • Collaborate with clients to create  entertaining and information based content marketing that provides value to reader.
  • Perform statistical and economic analysis on data sets to generate compelling  and engaging insights.
  • Synthesize results into a data-driven story that is easily digestible for an audience of a variety of technical levels.

Education
Georgia Institute of Technology
Master of Science Computer Science - Machine Learning
Expected 2018
The College of New Jersey
Bachelor of Arts Mathematics
2011, GPA 3.974/4.000
Metis Data Science Bootcamp
Certificate Data Science
2015
Credentials & Awards
Associate of the Society of Actuaries
Feb 2014
Society of Actuaries
Secor Alumni Award for Excellence in Mathematics
May 2011
The College of New Jersey
Robert S. Duncan Memorial Award
May 2010
The College of New Jersey
Skills
Math, Stats and Machine Learning
Probability Distributions
Bayesian Inference
Stochastic Modeling
Maximum Likelihood Estimation
Hypothesis Testing
Time Series Analysis
Generalized Linear Models
CART
Support Vector Machines
Ensemble Methods
Unsupervised Clustering Algorithms
Dimension Reduction
NLP/Topic Modeling
Tools/Technology
python
R
SQL
mongoDB
d3.js
flask
Excel
VBA
HTML/CSS
Git/Github
Familiarity with Big Data Technologies (Hadoop, MapReduce, Spark, AWS)
Projects
Digital Content Decision Engine

  • Utilized flask and d3.js to develop interactive web application for writers or advertisers to predict social media shares of digital content from Mashable.com and recommend ways to increase reach.
  • Engineered features from headlines and content using NLTK/Textblob/textstat and Latent Derelict Allocation to use in Poisson Regression and Random Forest Models.

Diagnostic Analytics for Health Benefit Elections

  • Analyzed how employees' interactions with benefit information distribution channels (web portal and call center) impact election decisions using Chi-Square Testing and Logistic Regression in R.
  • Visualized results with pie and ribbon charts created with ggplot2.

NBA Beat Writer Dashboard

  • Built interactive dashboard to aid casual NBA fans in selecting which NBA beat writers to follow on Twitter using NodeJS, D3.js, DC.js and MongoDB. 
  • Performed sentiment analysis with NLTK/Textblob and topic modeling with K-Means clustering.

Diagnostic Analytics for College Completion Rates

  • Analyzed key features affecting college completion using a Random Forest Model.
  • Created scatter plot using D3.js to be used for diagnostic analytics by college and public administrators.

TV Adaptation Analysis

  • Built model using regression analysis to predict financial success of movie adaptation of television shows from data obtained from the web using Beautiful Soup.
  • Implemented lasso regularization to extract key predictors of success.

MTA Traffic Analysis

  • Analyzed MTA data and prototyped dashboard tool to optimize deployment of street marketing teams in relation to target audience.