Eight key skills of a successful Data Scientist

Data Science   |   
Published October 26, 2015   |   

With the rise of Big Data in the past few years, a new breed of analysts has gained massive exposure. The so-called “Data Scientists” are considered by many as the only ones capable of leveraging the true value of Big Data. Although it has now become clearer what to expect from them, it is not fully clear what type of profile they should have. The most common mistakes being to think that a statistician is a Data Scientist or that a BI expert with some analytical background will be a good Data Scientist. Maybe in some cases, they will, but it will be the exception to the rule.

Key skills to become a true Data Scientist

1. Statistical knowledge:

It’s not only about knowing statistics, but it’s also about knowing the models and methodologies and how to best apply them. It’s about being a statistician with a strong set of analytical skills and knowledge of data analytic tools such as R, Python, etc.

2. Business Intelligence expertise:

It is not necessary to be a BI expert, but a DS needs to know how to extract, clean, transform, analyze and report data. A DS isn’t an ETL developer, but he/ she can speak with him and understand his jargon and activities.

3. Business analysis:

A DS isn’t expected to be a business analyst, but he’s expected to behave like one. The scoping of a case/ project requires in-depth discussions with business users to understand what they need. A DS should be included in the early stages of these discussions to rapidly grasp the concepts and participate actively in the definition of the deliverables.

4. Visualization:

Results of Advanced analytics often can’t fit in traditional visualization. A DS is expected to be knowledgeable with regards to the various visualization tools and techniques available.

5. Communication:

Communicating on the results and progress of the analysis is as important as getting those results. The process around Advanced analytics is often very iterative. Once first results are found, they should be communicated proactively to the users so that their comments are retrofitted in the next steps of the analysis. The DS is expected to fuel that iterative process. The DS is also expected to present the results of the analysis. For that purpose, a sense of storytelling is required.

6. Curiosity:

For the business/ topic under analysis as the DS might not be an expert in this specific field. Curiosity also on the evolution related to Data Science techniques, tools and technologies.

7. Creativity:

A DS needs to use a variety of tools and techniques in all the skills mentioned above. To do so, a DS should show creativity in how to use these tools and techniques. Knowing the case under analysis, he/ she should come up with innovative proposals as per how to analyze a specific issue, how to present a set of data, etc.

8. Common sense:

Last but not least, common sense should be applied wisely for the activities performed by a DS. If a case simply requires a linear regression and a bar chart, the DS should not over-engineer it. What is truly expected from a DS is to provide insight and added value information, not to reinvent the world.
This list is most certainly not exhaustive and doesn’t intend to be, but it does show that Data Scientists won’t be easy to find as the scope of skills they should possess is quite large and doesn’t correspond to previously known functions.
Please feel free to comment so that we can enrich and amend this list to come up with the most appropriate description. Building on your comments, I’ll publish a new post with an updated version of this one.
This article originally appeared here. Republished with permission. Submit your copyright complaints here.