What is Full Stack Data Science?

Welcome to our 1st Full Stack Data Science meetup located in Tysons. We are excited to have you join us for three great speakers, Mr. John Eberhardt, Dr. Anamaria Berea, and Almost-Dr. Kris Wright.

Agenda

6:00 - 7:00 - Networking

7:00 – 7:10 – Introductions and Sponsors

7:10 – 7:30 – John Eberhardt

7:30 – 7:50 – Dr. Anamaria Berea

7:50 – 8:10 – Kris Wright

8:10 – 8:30 – Additional Q&A / Discussion

8:30 – Data Drinks

Speakers and Abstracts

http://photos2.meetupstatic.com/photos/event/d/b/a/2/600_454556226.jpeg

Mr. John Eberhardt, ATA CTO and self-proclaimed oysterphile, will discuss “What the Hell is Full Stack Data Science? Why you can’t do Analytics in a Vacuum.” John will present lessons learned on how to increase the likelihood of data science project success by focusing on what happens before and after the data is curated and analyzed. Illustrated with colorful examples of success and failure – because the smart learn from their mistakes and the wise learn from the mistakes of others.

Dr. Anamaria Berea, a double PhD and avid tennis player, will be speaking about extracting meaning from emergent information. While exploratory data analysis and complex data science techniques can reveal very interesting patterns and new information, the interpretation and the understanding of these findings is usually left to the subject matter experts to turn it into actions. But this way of doing data science can be lengthy and expensive often leaving counterintuitive but valid data relationships unexplored. Thus one of the problems in data science can often be the divorce between the complex analysis and the context of the problem to be solved, particularly for new, fast data, noisy data.

Almost-Dr. Kris Wright will close with the problem of clustering in the online setting when the number of clusters is not known a priori, and when the clusters are non-stationary. He will demonstrate his new algorithm dubbed K-means+- due to the fact that the ground truth number of clusters could increase or decrease at any time. Kris will introduce several innovations to overcome the limitations of the classic k-means algorithm. Kris currently designs and implements platforms and analytics for streaming data and completing his studies at Old Dominion University where he is concentrating on the study of social network analysis and reinforcement learning.

Logistical Notes:

The space is appointed with plenty of seating, A/V, food, and (of course) beer. The conversation will begin at 7pm with about an hour of presentations. Tysons Biergarten is located at 8346B Leesburg Pike on the Silver Line at the Greensboro Metro Station. There is also ample, free parking.