Evan Kepner
MIDS 2016
Data Visualization and Discovery Consultant
Slalom Consulting
Education
B.A., neuroscience, University of Pittsburgh (focus on computational biology)
I School focus
I focused on natural language processing and computer vision. Classification problems like sentiment analysis and image recognition have helped push my boundaries. It's a cool way to mix in my love of language and art with data science.
An information challenge that intrigues you
Sensors are going to be ubiquitous, so how do we derive insight from billions of simultaneous data streams? Amazing things are happening with sensors and the scale is going to be huge. We think of sensor applications in personalized medicine or precision ag, but everything is going to be impacted.
Favorite class
Storing and Retrieving Data would be my top pick. That class was my “Yup, I’m a data scientist now” moment.
We built a system to mine Twitter sentiment around the UEFA championships, tied to geo-location and the game score, minute by minute. This included a custom search interface for classified text, interactive visualizations, as well as a graph database representing the relationships between Twitter users.
I felt like I could do anything after finishing that class.
The Berkeley I School advantage
The I School has an outstanding curriculum for data science and has provided the flexibility to keep material current. The curriculum wouldn’t matter much without strong execution. I was continually impressed with caliber of the faculty and staff. The online format was great; it made balancing the degree with work much easier.
Why did you choose the I School?
I chose the I School for the curriculum and its reputation. It was my first and only choice for a graduate program. Seeing Field Experiments and Ethics listed as classes was a clear indicator that this was more than just learning technology. As I read more about the faculty research it was obvious the school took cross-disciplinary focus seriously.
The best thing about the I School
The best thing about the I School is the people. The faculty and staff are fantastic, and the other students are some of the most interesting people I’ve ever met. Even though we’re distributed, we keep in touch and collaborate all the time. There is a real sense of community.
Advice for aspiring data scientists
Intuition and good science are key, and in many cases you don’t need to be that fancy. There’s a great paper called “The Unreasonable Effectiveness of Data” that drives the point home. A simple model with more data was better fit than a complex algorithm.