The hardest part of hiring is hiring the latest hot professionals, just as you were hiring silicon semiconductor experts in 1954 when the transistors were not available; or hiring HTML experts in 1996, when SGML experts were not in the government department It is at the publisher.
And in the last two years, "data scientists" have become one of the many companies that companies must hire, and to myself, in order to create Bright Score, a business lab that provides interesting data to the media and the public, The hiring team includes a former nuclear physicist, neurologist, geographer, astrophysicist and a mechanical engineer.
None of them have any working experience in data analysis, but they spent many hours researching, experimenting, analyzing and solving real-world problems. And their work has withstood the harsh criticism of peers and successfully published in academic journals, of course, this is my own company to recruit data analysts routines, then you should follow what guidelines?
First, what is a data analyst?
A Hadoop expert, a lot of hiring managers need such an expert; a machine learning expert; a Kagglers account
I personally define a data analyst as a candidate who understands programming, systems management, how unstructured data becomes operational, and how data is actually made. And he still has enough visualization and communication skills to let others know what he is doing. It is currently found that searching LinkedIn for "Data Scientist." Is the least effective way because data science is an emerging area where you want to find a person who has 5 years experience in data analysis does not exist.
The well-known bank robber Willie Sutton once said, "I rob the bank because I know there's money." And if you want to find a data analysis scientist, you can go to academia, doctoral and postdoctoral institutions to see, because they are in order To catch up with academic changes on the New Moon and the moon, maybe twice as much as the average person, in order to prevent the misuse of academic lazy bugs, you can ask them some questions before deciding whether to hire him:
1) "Can you cite some of the papers you published as first author?"
We need people who can do lengthy and complex tasks. It takes a lot of thought to publish a thesis that has been criticized by peers because colleagues can hurry and criticize you for blatant hiding behind anonymous walls.
2) "Cite some code you are writing to others?"
There are many academic theories but we need practical theory. The best way to test a code is to see if more than two people are using it.
3) "Cite some of the statistical methods you used in your essay?"
The best way to understand statistical concepts from a basic level is that everyone can list a series of ANOVA in the Excel spreadsheet, but it is not the best method, the data is not stuck, and the result is straightforward. .
The culminating analysis found in various journals or drafts is more than freelance scholars, the former being cheaper and more knowledgeable.
Via blog.bright
(Editor: left good)