Data Science in the Rough

Eli Bressert DEA @ Netflix

Data Sciencein the Rough

astrophysics

data by storm

123

Research & Academia

Application in Industry

Data Matters

universal law: EDA

Anscombe's quartet

select country, count(*) as frequencyfrom some_tablegroup by countryorder by count(*) desc

country | frequency-------------------NL 3US 2NZ 2MX 1

universal tool: ???????

universal tool: division- Monica Rogati

universereal world

1.3 billion light years away

in a far away galaxy

50 x more power than all the visible light in the Universe

WWhat do we have in common in this room with gravitational waves?

123

Research & Academia

Application in Industry

Data Matters

source: http://matt.might.net

Bressert et al., 2012

natural language processing

king - man + women =

king - man + women = queen

Paris - France + Italy =

Paris - France + Italy = Rome

computer vision

Computer vision examples

source: https://www.nextrembrandt.com

Computer vision examples

source: https://www.nextrembrandt.com

http://arxiv.org/pdf/1508.0657

strategy

traveling salesman problem

traveling salesman problem

123

Research & Academia

Application in Industry

Data Matters

A decade in academia taught me a bunch of

sophisticated algorithms; a decade in industry taught me when not to use them.

- Monica Rogati

did not use the top performing algorithm

result:

a/b testing

@ Netflix

2011

2013

data science isn't about the tools, it's about how you use them as a means to an end

123

Research & Academia

Application in Industry

Data Matters

data storage

Data Moats- Pete Skomoroch

start banking questions

nearly done

the future of data science?

Data science in the future

Data science in the future

all things data will be ubiquitous

imagination is your only limit

@astrobiased @netflixdata [email protected]?