Follow Datanami:
August 22, 2019

Skills Are Critical in Data Science Job Hunt

Alex Woodie

(SK Design/Shutterstock)

Those planning a career in data science have a healthy job outlook, as demand for data scientists continues to grow. While an advanced data science degree can definitely help, it’s becoming increasingly apparent that having the right skills is a more critical factor in landing your dream job.

Demand for data scientists continues to outpace supply, which is a good thing if you’re on the labor side of the equation (and not so good if you’re management). Data scientist is the number one job in the country, according to Glassdoor’s 2019 rankings, with a median base salary of $108,000 and more than 6,500 openings.

According to data from the Bureau of Labor Statistics, folks who are employed in the field of “computer and information research scientists” (the closest thing to a “data scientist” in BLS’ catalog of jobs) earn an average of $118,370 per year, or nearly $57 per hour. The number of jobs for these computer and information research scientists was projected to grow nearly 20% from 2016 to 2026, the BLS says.

A master’s degree has traditionally been thought of as the key to landing a data science jobs. For example, the BLS says a Master’s degree is typically required for research scientist positions, which is unlike other jobs in the computer industry, where BLS says Bachelor degrees are typically sufficient. Having a PhD in data science (or similar fields) has been an even better indicator of future success.

If you’re coming out of a university with a data science degree, you most likely have a master’s or a Ph.D., as undergraduate data science programs are still ramping up. Coming from these programs, you will undoubtedly have taken courses in statistics, mathematics, machine learning, computer science, and programming. If your data science education came within a computer science department, you probably have familiarity with distributed and parallel systems. These are fundamental building blocks for data scientists, and form the foundation for their work.

The degree will prove that you’ve mastered these core concepts required to be a data scientist. However, as the pace of innovation in data science has accelerated, degrees have lost a little bit of their punch in the job market. Employers are now looking for data scientists with a certain set of skills, and it doesn’t necessarily matter how you got those skills.

Take the case of David Venturi, who dropped out of a graduate-level computer science program to teach himself data science. In his 2016 Medium post, Venturi says he was able to glean necessary skills by taking online classes by the likes of Udacity, edX, and Coursera.

“I could learn the content I wanted to faster, more efficiently, and for a fraction of the cost,” he wrote. “I already had a university degree and, perhaps more importantly, I already had the university experience. Paying $30K+ to go back to school seemed irresponsible.”

So what are the skills that you need?

Some of the languages that are currently in high demand include Python, R, Java, Scala, SQL and MATLAB. Many businesses have invested in SAS, which like MATLAB is proprietary code backed by a single company. But while MATLAB can be fined in academia (particularly engineering departments), it’s rare to run into SAS in college.

When it comes to machine learning frameworks PyTorch, Tensorflow, Caffe2, scikit-learn, and Keras arguably are the most popular at the moment. The do-it-all Apache Spark project is used by data scientists and data engineers alike, and is largely replacing Apache Hadoop as the core platform for big data.

In the ivory tower version of the data science profession, data scientists spend their days working with the latest algorithms (random forest, K Nearest Neighbor, or neural network) to extract the highest predictive power possible. In the real world, data scientists spend the bulk of their time working with data – extracting data, moving data, loading data, working with databases and file systems, and cleaning and prepping data – and only a fraction of time is spent on “real” data science work.

To that end, it behooves the would-be data scientist to also bolster their resume with some data engineering skills. You should learn the concepts around extract, transform, and load (ETL), or its close cousin, ETL. You should know how to write a SQL query, how to work with common databases like Postgres and MySQL, and how to work with cloud services, such as AWS S3, Azure ALDS, and Google Cloud.

You will probably run into big, on-prem data clusters in your career, so knowing your way around Hadoop won’t hurt. Many companies are using NoSQL data stores like MongoDB and Apache Cassandra to hold semi-structured data, like JSON, while more specialized NoSQL stores called graph databases (Neo4j and AWS Neptune) are used for finding links among entities.

You can’t rely on algorithms to tell you about data all the time. Sometimes your eyes are the best tools, so finding ways to visualize your data sets is important. D3.js arguably is the most popular open source visualization tool, but if you’re willing to shell out some cash, other alternatives include Tableau, Google Charts, Domo, Microsoft PowerBI, and FusionCharts.

With unemployment low and demand high, now is a great time to be looking for work in data science. By keeping your skills up to date, you’ll maximize your earning potential and have companies banging down your door to hire you.

Related Items:

Data Science Back to School: Accelerate Your Education

Is Python Strangling R to Death?

 

Do NOT follow this link or you will be banned from the site!
Share This