Ferris Jumah, a data scientist at LinkedIn, published an article on the network about the difficulty of describing the characteristics of his profession, even though there are already several materials offering descriptions and guides about this career. Jumah did something different: he analyzed data extracted from the LinkedIn platform itself, based on the profiles of people who called themselves “data scientists”. Thus, he was able to take a more assertive approach to defining this professional.

Most popular skills among data scientists, according to LinkedIn

Data Mining

Machine Learning

R

Python

Data Analysis

Statistics

SQL

Java

Matlab

Algorithms

Although this list shows us which skills are most commonly found in professional profiles, it is difficult to relate them by looking at a ranking alone. To make it easier to understand, Ferris Jumah went deeper and created the following chart:

(click here to view the image in a higher resolution)

Each node represents a skill. To facilitate the visualization and understanding of this information, they were grouped by similarity and represented by colors. Then, they were scaled according to how many times they appeared connected and their influence on other skills in the network.

Several conclusions can be drawn from this study. Here are some of them:

1) Approaching data with a mathematical mindset

We see that machine learning, data mining, data analysis, and statistics have a high ranking. This indicates that being able to understand and represent data mathematically, with statistical intuition, is a fundamental skill for data scientists.

2) Use of a common language for data access, exploration, and modeling

Python, R, and Matlab are the three most popular languages for visualization and development modeling, and SQL is the most common for data access. When it comes to data, extracting, exploring, and testing hypotheses is a big part of the job. It is no surprise that these skills are in the spotlight.

3) Strong developers

We also see computer science and software engineering as qualifications, along with Java, C++, Algorithms, and Hadoop — all having a notable space in the network visualization. These are the skills mainly used to leverage data for systems development.

There is probably no single professional who is an expert in all of these skills, but rather in one or two of them. This is, therefore, a holistic view of the characteristics represented within a typical data science team.

And you, did you reach any other conclusion from Ferris Jumah's study? Share it with us!