Full Stack Data Science #2


Details
The agenda for our next meet-up is now set. Our final agenda and speakers for the February 7th event.
In addition to Patrick Smith and his talk on Scaling Deep Learning, we are glad to welcome Dr. Jonathan Seale who will be discussing the Case for Ontologies. We are also introducing a new "Lighting Talk" series with an inaugural presentation by Mike Gallaher.
http://photos2.meetupstatic.com/photos/event/8/5/5/9/600_458014137.jpeg
Here is the agenda for the evening:
Agenda
6:00 - 7:00 – Arrival, Networking (w/food and beverage)
7:00 – 7:15 – Introductions and Sponsors
7:15 – 7:45 – Patrick Smith
7:45 – 7:55 – Mike Gallaher Lightening Talk
7:55 – 8:25 – Dr. Jonathan Seale
8:25 – 8:30 – Wrap-up / Additional Q&A / Discussion
8:30+ - Data Drinks / Networking
Speaker and Topics:
Patrick Smith will be presenting Deep Learning at Scale. Deep Learning has been ubiquitous in it’s advances in recent years. One of the hardest problems that we’ve all faced, however, is how do we scale it? Patrick Smith (brief bio below) will be talking about how to deploy deep learning algorithms at scale, and how to integrate these algorithms into your greater data science stack. More about Patrick: He is a data scientist and all around deep learning enthusiast currently working with Excella Consulting. When not attempting to replicate the human brain in code, he is an avid oenophile and antique collector. Patrick has worked in data science in the consulting and financial services industry, and previously was a data science instructor. He has an undergraduate degree from George Washington University and is currently pursuing graduate studies in AI remotely at Stanford University.
As I mentioned before, we are launching a new lightening talk series to make it easier for more members of the meet-up to participate and share. Lightening Talks are 5-10 min each and focus on one of the three following topics: Data, Data Everywhere!; I Think I Found Something New, Did I?; and Failure with a Capital F.
Our first lightening talk…Data, Data Everywhere!
You can find data in places you didn't expect, though sometimes you have to squint a little (and it doesn't have to be big to be interesting). In this talk, Mike Gallaher will show how sound can be extracted from a 2D image of an oscillogram trace printed in an old magazine article. This is an example of what Patrick Feaster calls "educed audio" (where audio is derived from representations of sounds that were not intended to be played, such as photographs of phonograph records). By day, Mike is a software generalist currently working as a DevOps lead in an agile software development shop. An EE by training, his first love is electronics. By night, he builds gadgets with his home built CNC router and tinkers with Software-Defined Radios.
Following Mike, we will welcome Dr. Jonathan Seale. He’ll make his Case for Ontologies. Ontologies are a logic-based structure for organizing knowledge. They define the terms within a knowledge domain and the relationships between those terms. The difficulties with sharing diverse, high-volume data can be alleviated by creating the common vocabularies defined within an ontology. This talk will outline the basics of what ontologies are, how they are constructed, and how their use can be beneficial in a data science environment. Finally, Jonathan will demonstrate some ontological tools and use-cases so that you can explore using ontologies in your own work. Jonathan is a data scientist and PhD astrophysicist currently working for ATA, LLC. Prior to working in industry, Jonathan was an astronomer at Johns Hopkins University and the Space Telescope Science Institute, where he used data science techniques to identify and classify forming stars. When he's not designing ontologies for the DoD, he teaches college-level astronomy and takes astronomical images from his home telescope.
Logistics:
Greenhouse Bistro is a great new, modern spot in Tysons (although the address says Vienna) located at 2070 Chain Bridge Rd in what is affectionately known as the Toilette Bowl Building. There is plenty of parking around back. Closest Metro spot is on the Silver Line - Greensboro Station. Its about a 5 min walk heading Southwest. Food is provided. Drinks on you.
It’s going to be another great night!

Full Stack Data Science #2