Data Scientist Team Share Lessons Learned From Competition

May 31, 2017

10:30 am

When you’re in a competition and on the spot, answering questions to industry experts can be a nerve-wrecking process. The best part is that afterward you walk away with valuable insight, experience and lessons learned.

GalvanizeU graduates Baolin Liu and Jonathan Hourany, current gU student Anthony Abercrombie, Data Scientist in Residence Eugene Huang, and their teammate Jared Wilber, a UC Berkeley grad student, learned a ton after their time on stage at the iidata Time Analysis Competition.

This competition required the team to answer “…questions on a Food and Nutrition data set that contained 28 features and 30,000+ rows. We were given 3 hours to solve five ‘expert level’ questions including: finding bias in nutrition scores, classifying food categories, and predicting originating country,” said Johnathan Hourany’s in his post.

The first time they competed didn’t go as well as expected and placed 2nd. They left the competition with their lessons learned and came back to win first place in 2017.

“To us, the ‘expert questions’ were like the typical machine learning questions we’d get in a Galvanize classroom,” said Baolin Liu.

lessons learned, data science, GalvanizeU, Galvanize

Here is what Liu and Huang did to tweak their approach to the 2017 event based on these lessons learned:

  • “Get to the church on time.” Last year, the guys were late to the competition because of major traffic delays. This time around, they allowed more commute time and arrived early.
  • Speaking of time: They were smarter about allocating enough tick-tock for each phase of the process leading up to their final presentation.
  • Put it in the slides: At the 2016 event, Liu & Huang overlooked the need to include some of their key findings in their slide deck. This time, they made sure to include all of their pertinent analysis in their presentation.
  • Delivery is key: The guys picked up the pace of their presentation after going too slowly in 2016. They made sure each team member knew his part well, and they timed themselves to be sure they hit the mark.
  • Test your skills: Create a mock competition environment and have experts or mentors in your field be the judges.

The G-team also had another year’s worth of education and experience under their belts, so their skills and confidence were that much stronger.

When it came to answering the judges’ questions, says Liu, “We got grilled by our Galvanize professors all the time, so [the iidata judges] didn’t seem so intimidating.”

Read more tips from founders on approaching competitions at Tech.Co

This article is courtesy of the Galvanize blog. Interested in entrepreneurship, web development, or data science? Interested in entrepreneurship, web development, or data science? Check out the Galvanize Newsletter, bringing you the best content from The Learning Community for Technology.

Did you like this article?

Get more delivered to your inbox just like it!

Sorry about that. Try these articles instead!

Galvanize campuses are home to an eclectic mix of entrepreneurs, skilled programmers, expert data scientists, and men and women motivated to make an impact. As different and diverse as we all are, we’re united by common values. Interested in entrepreneurship, web development, or data science? Check out the Galvanized Newsletter, bringing you the best content from The Learning Community for Technology.

  • Shares

Leave a Reply

  • (will not be published)
Startup_Mixology_300x250
.ME Tech.co search 300x600

Get personal with .ME

Make your domain name as unique and memorable as you are. Get creative, get .ME!
Startup_Mixology_300x250