How to Move From Math Major to Data Scientist

So, you have a degree in math and want to become a data scientist. How do you get there?

The path from math to data science isn’t often a linear one, but there are some steps you can take to get started down the right path.

Make Sure You Have the Right Degree

If you realized late in your bachelor’s degree, or even after you graduated, that you wanted to go into data science, you’re not alone. That’s pretty common, according to Heather Ramsey, lead undergraduate advisor at Texas A&M University’s Mathematics Department. Unfortunately, most math bachelor’s degrees that aren’t targeted to data science don’t include the necessary programming classes for a job in the field. So, you might want to consider pursuing a master’s degree or a doctorate. 

If you are specifically after the title of data scientist, you need to pursue a graduate degree. Most job postings for that title ask for advanced technical degrees.

If you opt to go for an advanced degree, check with your department to see what, if any, data science-focused programs are available there or look for a school that has one. While data science as an academic discipline is still relatively new, it is growing. Even if your school doesn’t have a dedicated data science program, more and more math departments have minors or interdisciplinary programs that partner with computer science or engineering departments to get students the necessary statistics, data analysis and programming classes they need.

More on Data ScienceHow to Land Your First Data Science Job

Figure Out What Element of Data Science Interests You Most

Data science is a broad field encompassing many sub-disciplines from AI and machine learning to business intelligence and data visualization. These all come with different titles and roles within companies. 

Knowing what you like doing — are you most interested in machine learning? Natural language processing? Solving security problems? — and where you want to do it will help you focus your skills as you work towards your career.

Ian Wong, co-founder and chief technology officer of Opendoor, a digital real estate platform, knew he wanted to build data technologies that could have an immediate impact. He learned that the cutting edge of data technology development was happening in industry rather than academia — so he left academia to pursue a career in data. At his first data science job at Square, he wound up building internal account review tools to protect against fraud.

Learn Outside of Class

One of the hallmarks of the data science field is the need to constantly learn and keep up with changing technologies. Whether you are continuing your education formally or not, you should explore learning resources.

Melinda Han Williams, chief data scientist at Dstillery, a custom audience solutions company, said aspiring data scientists with math degrees should learn the fundamentals of statistics and machine learning through books. While Stack Overflow and Medium are great for general data science information, perspectives and tips, they shouldn’t be your go-to resources for foundational learning, she said.

There is a plethora of books geared towards data science (or its many subdomains), including textbooks aimed at beginners. Jacky Koh, co-founder of Relevance AI, a vector platform for AI experimentation, named An Introduction to Statistical Learning as an important book in his path from his statistics major to his career as a data scientist. He said it helps readers develop foundational understanding of statistical learning and contemporary data collection practices.

David Dillmann, senior data scientist at Allianz, a German financial services company, also advised going beyond your coursework and suggested classes from Coursera and Udemy. One of the best parts of the data science field, according to Dillmann, is how many resources are freely available online to help improve skills and knowledge.

Koh suggested the Coursera machine learning course by Andrew Ng as a good resource for people with math backgrounds who are getting into data science.

Data science bootcamps are another good option. They offer rigorous, intensive learning experiences outside of the usual academic classroom.

Learn Python (at Least)

Unless you are pursuing an advanced degree targeted to data science that includes programming classes, part of that learning outside of the classroom should include learning Python. Even if you are getting Python instruction in a class setting, you should still consider augmenting it with external learning.

Not only is Python one of the most popular programming languages for data science, there are a lot of resources out there to help you learn it.

You may want to learn other popular data science languages as well, depending on what your area of data science interest is. R, SAS and SQL are good choices that show up in job postings frequently.

Get Some Real-World Experience 

The most critical step on the path from a math degree to a data science career is to get real-world experience.

“What [university] teaches students is still only a fraction of what is expected from a professional developer and data scientist,” said Dillmann. “Real-world data is messier than the cleaned up data sets you get presented with during class.” Working in data science is also more than working with data, he added. It includes business strategy, legal understanding, interacting with important stakeholders and effective communication.

“What [university] teaches students is still only a fraction of what is expected from a professional developer and data scientist.”

There are several ways to get real-world data science experience while you’re still in school or working on doing self-directed study. These include internships, fellowships and participating in open-source projects or competitions. 

Internships and Fellowships

Beyond experience, internships can also help you figure out what areas of data science most interest you from a hands-on perspective. For Williams, her internship experience was part of what inspired her to pursue machine learning, for example.

Anywhere you can look for job postings you can also find internships. Your university’s career center — more on that later — should also be able to help you find available opportunities.

Fellowships have a lot of the same benefits of internships. They can also come from a variety of sources, but are usually only available to doctoral or postdoctoral students, though some exceptions exist. Check with your university or department for school- or location-specific fellowships. 

While fellowships usually come from academia, they can be found elsewhere, too. Major tech companies like Google, Microsoft, IBM, Amazon and Meta all offer fellowships. Math and data science industry groups like the American Statisticians Association offer fellowships as well.

Projects and Competitions

Open-source projects and data science competitions are widely available regardless of where you are in your data science career path. They might even be the key to getting your first job. Briana Brownell, founder and CEO of Pure Strategy, an AI-driven analytics company, said her participation in COMAP’s Mathematical Contest in Modeling led to her first job in data science.

“Through the contest, I had to work on a small team and solve a problem over a weekend,” she said. “So I’d recommend students do something similar — find an opportunity to take on a project to prove you can solve real problems. Bonus points if you’re working with other people.”

Koh had a similar experience with the EY Data Science Challenge, which he won in 2018. Winning the competition earned him an internship with EY, which Koh described as getting him his start as a data scientist.

“Find an opportunity to take on a project to prove you can solve real problems. Bonus points if you’re working with other people.”

Kaggle competitions are a good way to get some real-world project experience. Google’s Summer of Code is another high-profile option. Math majors working towards a career in data science, however, should also try independent projects of their own creation that solve a real problem, Williams said.

“Resist the temptation to start with a data set and then think of what you can use it to predict,” she said. “A big part of data science is figuring out how to frame and address a business problem. Pick a domain, think of an actual problem someone might want to solve and then figure out what data and techniques you can use to solve that problem.”

Create a Portfolio

Those projects will fit nicely in your data science portfolio. If you don’t have one yet, make one, especially if you don’t have any data science work experience to show.

“Without having a portfolio of projects, you will only be able to demonstrate your understanding and knowledge, not your productive and tangible application of it,” Koh said.

Be strategic in what you include in your portfolio. It should include projects with real-world applications that show off the breadth of your skills and highlight your strengths. It helps to communicate about the process behind the solutions you came to in the projects. This can be done through project write-ups or blog posts. To stand out, give potential employers insight into how you think since so much of data science work is based on individual problem-solving.

More on Data Science10 Steps to Become a Data Scientist

Use Your University’s Career Services

Universities have a wealth of services to help students (and alumni) find internships, fellowships and jobs, as well as help with your career direction. 

The best first stop is your university’s career center, according to Ramsey. You can do this whether or not you are a current student. University career centers are there to help students and alumni alike find jobs and other career resources. Depending on your school, there might be different offerings or procedures to go through if you have already graduated. There may also be discipline-specific career counselors who will be more knowledgeable about math and data science fields and direct you better than more general career counselors. 

University (or departmental) career fairs are also often open to students and alumni alike. Check to see when these are — and, if you’re an alum, if they are open to you — and go to them if you can. 

Employers who present at university job fairs know they are going to be talking with people who are students or otherwise don’t have much work experience. Talking with these employers can be a great way to find out who is hiring at your level and what they are looking for in terms of skills and experience. If you are still in school, visiting career fairs can help you tailor your later classes and target your degree, according to Ramsey.

Other job-related resources to look into would be your university’s job board or see if your department has something similar that is more targeted to math.

Practice Your Data Science Skills

As you learn and progress down your career path from your math degree to data science, be sure to practice the skills you learn. Ramsey stressed the importance of practicing with programming, in particular. She has seen some math undergraduates who go on to pursue data science discover that they don’t enjoy programming as much as they thought they would. Since programming is essential to most data science roles in addition to a strong math foundation, make sure you like doing it.

“Your skills as a data scientist are completely dependent on how willing you are to practice.”

Albar Wahab, associate data scientist at Data Science Dojo, a data science training platform, also emphasized the importance of practicing — and just playing with data and problem-solving on your own time. There are plenty of free data sets that can help you learn, explore and practice your new data science skills.

“Your skills as a data scientist are completely dependent on how willing you are to practice,” Wahab said. “There might be some stages that you will find cumbersome, such as cleaning and preprocessing the data, but you need to remember that it is part of the process. The edge that you have with a mathematics degree is that behind all the machine learning algorithms, analysis and finding insights is an integral part of data science, which is mathematics.”