Well, my journey to becoming a data scientist was not an intentional journey - I didn't start with a plan to end up in a data science career. In fact, I ended up doing data science for years before I realized it had a name.
What is a data scientist?
Data scientists have been described as a mashup between mathematicians and computer scientists with a curiosity to solve unexplored problems. There's plenty of data scientist job descriptions out there, but I think this description sums it up. If you'd like to learn more about what it means to be a data scientist, I recommend this article.
They’re part mathematician, part computer scientist and part trend-spotter.
Where did I start? Math.
If I were to try to pinpoint the start of my journey, I would have to say it started with a love of math. I started college as a mathematics major and briefly had the idea that I was going to become a mathematician.
My math program required an intro to computer science class, which I took my first semester and met my eventual undergraduate advisor. As part of the class, she met with each person individually to discuss their goals for the class and their career. In that meeting, she showed me the job opportunities (and salaries) for a mathematician versus a computer scientist. Let's just say that computer science looked more promising and as someone who always wanted to say they were a scientist, I was sold.
Shortly after that class, I started doing research under her. I was using machine learning to predict fraudulent energy consumption usage, which continued until I graduated and led me to apply for Ph.D. programs. At that point, I had already been doing and loving data science, but I wasn't aware of it yet.
I applied to several grad schools and decided to seek out labs that worked with educational data. Education provided me with many opportunities, and I wanted to be able to contribute to research that made education more accessible and attainable for others. After getting acceptance and rejection letters, I reached out to a research professor at one of the schools I was accepted to whose research focused on educational data. I met with her and made my choice to join their 5-year Ph.D. program.
You may be wondering at this point if you need a Ph.D. to become a data scientist and the answer is, as always, it depends. Data science is a quickly expanding career and multiple levels of data scientists exist. Some positions may require a Ph.D., but you will find many more that are just looking for a particular skillset and analytical experience.
Academia or industry?
During the program, I learned experimental design and setup, advanced analytics techniques and more about machine learning. After finishing coursework requirements in the first two years, I spent most of my time doing data analytics and writing up the results for academic papers.
Here, I also discovered a love of teaching through being an instructor. I dreamed up a new idea of becoming a research professor, but after I witnessed the day-to-day lives of professors, I started to doubt that I would enjoy it as a career. So, industry it was.
Luckily, I found that the opportunity to teach extends well past the academic world - after all, people want to learn about data science. In my current job at SAS within the Education department, I teach and design courses for our machine learning and analytical tools, although I spend most of my time doing typical data science work.
I use virtually all the skills I learned from my Ph.D. program, including skills I learned working with educational data. On a day-to-day basis, I work on designing, monitoring, and analyzing experiments to address business concerns. I also spend a considerable amount of time doing a variety of data processing tasks from cleaning to feature engineering, as well as visualization of data to help inform and improve internal processes in the SAS Education team.
What are some challenges of transitioning from academia to industry?
New domain, new audience, and most importantly, new goals!
Transitioning from academia to industry was, for me, a relatively easy process; however, there were a few challenges.
The biggest challenge I had was changing my mindset when it came to analytical goals. Compared to academia where the analytical process you follow to answer a question is just as important as the goal (answering the research question), the goal tends to be the focus in industry. For me, this meant updating how I communicate results and their implications. I also needed to learn an entirely new domain (business operations) and how to communicate results to a non-technical audience.
Understanding the needs of your company is the most difficult challenge by far, so asking questions is key! Presenting your interpretations or concerns of the goal is crucial to doing the job right. If you are someone in the process of transitioning from the academic world to industry, be ready to adapt to new processes and shift your mindset to accommodate solving business challenges, as the goals in business will differ from the goals of academic research.
Interested in becoming a data scientist?
If you are interested in becoming a data scientist, ask yourself:
- "Am I the type of person who digs deep into the problem to find the answer - often generating additional questions and answers in the process?" or
- "Would I rather focus on more defined, straightforward issues?"
Feel like you fall into the first type? Check out SAS Academy for Data Science (free for 30 days).
International Women's Day - March 8
In honor of International Women's day coming up March 8, I would like to thank all of the powerful women that have supported my journey.
I’ve met many inspiring women along the way. Both my undergraduate and graduate advisors happened to be women as well as my current boss, and I’m incredibly grateful to have had these leaders to look up to.