10 skills employers need in a data scientist

10 skills employers need in a data scientist

At this point, nearly every enterprise is looking for skilled data scientists to wade through buckets of information and glean actionable business insights. However, a shortage of professionals in the field, and the unique combination of skills the role requires, makes finding the right hire difficult, said Pavel Dmitriev, former principal data scientist for analysis and experimentation at Microsoft, and current vice president of data science at Outreach.

“It’s extremely hard to try to find and hire good data scientists,” Dmitriev said. One reason for this is that data scientists require a combination of hard and soft skills to succeed, he added.

There is also a change underway in the data scientist role:The vision of a data scientist is shifting from someone who is given information and a problem, to someone who understands the most pressing business problems that can be solved with data, and then can go collect that data and work with other teams throughout the company to drive value, Dmitriev said.S

“The qualities and skills data scientists should have are not limited to just coding and algorithms—it also requires an ability to tackle those problems,” Dmitriev said.

The role has also become increasingly broad, Dmitriev said, so it’s important for employers to be very specific in terms of the skillsets they are looking for in a job posting.

Here are 10 skills every data science hire should have, according to Dmitriev:

1. Coding

2. Algorithms

3. Big data analysis

4. Data manipulation

5. Statistics

6. Machine learning

7. Natural language processing

8. Exploratory data analysis

9. Formalizing problems

10. Communication

Natural language processing and machine learning skills may be optional for some roles, but more and more are requiring these abilities, Dmitriev said.

Meanwhile, exploratory data analysis involves receiving the data, understanding what is happening and what interesting insights are there, he added. Formalizing problems means taking a business problem and translating it into an actual mathematical data science problem that can be solved, which data scientists are also responsible for.

Communication is a vital skill, as data scientists must work with a number of different stakeholders and teams, and must explain complex mathematical concepts in a clear way, Dmitriev said. “An ability to communicate with everyone and explain what is happening is very important,” he added.

This and other soft skills are more difficult to test in a job candidate, Dmitriev said, so companies may want to give them a business problem to complete during the interview or at home that shows their analytical skills.

Source:

By |2018-12-14T13:58:44+00:00December 14th, 2018|Uncategorized|0 Comments

About the Author:

DataField