John Rauser: Classical analytical statistics relies on assumptions that are rarely true
John Rauser · John Rauser, Pinterest // Big Data at Pinterest // Data Driven NYC (Hosted by FirstMark Capital) · Oct 16, 2014 · at 12:24
John Rauser, Lead Data Scientist at Pinterest, compares computational simulation and resampling techniques against classical analytical statistical approximations during a Data Driven NYC presentation.
0:00 / 0:10exact quote · 10.9s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“The analytical approach is not better. This is an approximation to a distribution under a model with a whole bunch of assumptions which are rarely true and are almost always ignored anyway.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from John Rauser
Pinterest data scientist John Rauser cannot derive the t-distribution from scratch
“There might be five people in this room, the people who raised their hands, who could sit down and just derive this from first principles. I am certainly not among those people, and I am a working data scientist.”
John Rauser says many software engineers fake their understanding of statistics
“What I want to claim today is that many of you are faking it when it comes to statistics.”
John Rauser: Working data scientists must know how to code
“Yeah, I mean, I think, ah, like a working data scientist has to be able to code, because if they can't fish for themselves, right?”
A study shows beer drinkers attract 4.4 more mosquitoes than water drinkers
“To get the, to find that the average person who drank beer attracted 4.4 more mosquitoes than the average water drinker.”
John Rauser says programming gives developers superpowers for learning statistics
“If you can program a computer You have superpowers when it comes to learning statistics, because being able to program allows you to tinker with the most fundamental ideas in statistics, the way you might have tinkered with electronics when you were a kid, or …”