When I began my career in the analytics space in 2011 it seemed important to me to gain specific, specialized knowledge in certain areas. I had previously been a “jack-of-all-trades, master of none” kind of guy and that was not where I wanted to be professionally. Business analytics became that area of focus but after eight years I know now that once you gain a certain competency in data analytics you realize that they are innumerable subdomains; I am now, once again, somewhat of a jack-of-all-trades. Gaining mastery in the subdomains of data science require extensive practice and specific technical skills. Working with large companies I have generally found that my level of technical expertise across analytical methods is usually sufficient and it is of great benefit to know a little bit about a lot of different things.
In my projects I am typically functioning as the data architect, project manager, developer, scientist, QA tester, BI analyst and liaison to the customer. I have cut my teeth and learned through a lot of mistakes and now am skilled at taking a project from the point where it’s just an idea to a finished product, with responsibility for each step along the way.
I am currently working on a contract basis as a data scientist with Consumers Energy, a gas and electric utility headquartered in Jackson, Michigan. I work with their Customer Operations team. From 2012 - Sept 2018 I worked as an analyst at Jackson National Life.
The ML/Statistics techniques I’ve used the most include: anomaly detection, force-directed network visualization, logistic regression, time series forecasting with exponential smoothing and unobserved component models, automatic dimension reduction and variable selection, text mining, classification and regression trees, fuzzy matching, naive bayes and k-modes classification.
For analytics work I primarily work in SAS, R and SQL Server. Below are the packages I use most commonly.
SAS: Data steps, SQL, macros, SAS/GRAPH, SAS/STAT, SAS/ETS.
R packages: RMarkdown, ggplot2, tm, corrplot, changepoint, tidyverse, simmer, forecast, flexdashboard, rpart, RODBC and more.
Prior to Jackson I earned a master’s in Economics from Central Michigan (2010-2012) and worked as an intern for the Business Analytics team within Kellogg’s Sales department. I did research while at CMU on the elasticity of public transportation prices with a health economics twist. My master’s paper Fare or Foul: The effects of public transit fare increases on auto fatalities was the result of that research. This is the abstract:
A municipal transit authority considering a fare increase may forecast some of the financial effects of its policy change on its riders. But it is possible that price-setting should take into account other welfare effects on non-riders, such as public health consequences from increased automobile crashes. In this paper I examine the relationship between bus fares and traffic fatalities for 168 transit service areas in the United States from 2000-2009. Fares are significantly and positively related to total fatalities with an even stronger association with fatalities of children. Using two cities in the sample, New York and Des Moines, I create scenarios to forecast the increased fatalities associated with a 50 cent fare. New York would experience 20 extra fatalities, six from children, while the smaller Des Moines would experience on average one extra fatality with a 50% probability of it being a child.
I completed my undergraduate studies at the University of North Carolina at Chapel Hill in 2006. For one year I worked for a non profit in Champaign, IL as an Americorps VISTA. Afterwards I did some microfinance research in India and then worked a variety of jobs in the Champaign area.