Tech IndustryOct 26, 2018
CiscoijGA87

Python, R, SAS, or Excel/Microsoft Bi for Datascience and Business analytics?

A new hire I have thinks he's a data scientist. He's really data wannabe. What should I have him get really good at, Python, R, SAS, or Excel/PowerBI? Note that I'll probably have to get up to speed, I'm in a non-tech business role, so I can double and triple check his work....thoughts?

Amazon Yhbvcghj Oct 26, 2018

Sounds like you need to be less cocky.

Microsoft AXTc72 Oct 26, 2018

I mean what are the business needs? You canโ€™t just be like hey learn this deep learning library in Python if heโ€™s just going to be filtering excel spreadsheets. Also, donโ€™t be a dick - heโ€™s just started and youโ€™re already trying to degrade him.

Oracle of ยท Nada Oct 26, 2018

The tech doesn't matter, the solution does. If you insist, Python and(!) Java/C++ Find several teams with big data, and come up with a use case for something that mixes them. Have the new hire work on the problem start to finish, and ask that at least two recent papers be incorporated in the modeling phase. For end results, ask for materials for pitching to stakeholders and for results accessible in production (pipeline that takes new raw data, scores, and uploads everything important like scores and metrics for tracking model lifecycle to your DB) If you want to make it artificially cool sounding, force this paper about a quantum speed classical algorithm written by a child prodigy into it ๐Ÿ˜‚ heads up though, terrible practice to force specific solutions just to sound cool https://arxiv.org/abs/1807.04271

Oracle of ยท Nada Oct 26, 2018

And please find someone who knows what they're talking about do some of the checking with you...

Qualcomm Blahblah13 Oct 26, 2018

Data wannabe? Someone needs to cut you down to size. He/she youโ€™re talking about should run.

Cisco Sufgdd Oct 26, 2018

He/she is fresh out of college and knows how to do a VLOOKUP....talking to me about Normal Distributions like they know how to program in SAS and STATA So yes a data Wannabe. I'm very ok with it, knew what I was getting, but want to nurture the drive. Those who run are only running from their weaknesses....something you probably struggle with? Always running?

LinkedIn Gill Bates Oct 26, 2018

Tell him that most real world data in the tech industry isn't normally distributed. Latency isn't normally distributed, service usage isn't normally distributed, etc.