While I think that you’re generally going to be better off with a college degree than attending a boot camp, it might be useful to ask when might completing a data science boot camp be useful. The answer largely depends on your background, experience, and how you learn best.
I haven’t attended a boot camp, so I’m basing my assessment of data science boot camps on their posted curriculum and speaking with boot camp graduates. Take this answer with a grain of salt. From what I understand, I would group the role of boot camps into two groups: as introductory training to data science and as lateral transitions to data science.
Many data science boot camps do not have requirements or have minimal ones related to programming or statistics. Because these boot camps are aimed at a broader audience, the topics covered and skills developed will necessarily start from the bottom. The boot camp curricula that I have seen mix in presentations or lectures with hands-on application through projects, typically with a larger, end-to-end final project.
The structure and curriculum of many data science boot camps actually lend themselves well to learning the material and developing skills in a short period of time, thanks to the following features:
- Applying learning in projects
- Feedback during learning
- Working with real-world data
What makes it introductory training is that given the shorter duration of boot camps as compared to college degrees, the trade-off you’re likely to see is in the depth of data science knowledge, breadth of general knowledge, and less experience from having completed fewer projects. While boot camp graduates can get a good foundation and can probably start producing value as a data scientist after graduating, they just haven’t had the time to gain the experience and deeper knowledge that would help them identify and work through the problems with data and models that weren’t seen during boot camp.
The other element of a college degree that distinguish it from the vocational training of data science boot camps is the exposure to other disciplines that give context to data science and hone critical thinking. Data don’t exist in a vacuum and doing data science is more than optimizing loss functions. Concerns about algorithmic bias can be addressed through a background in sociology, ethics, political science, and even a better understanding of research methods. Normative questions on the use and effects of machine learning and AI in society can benefit from a background in philosophy. Further studies in mathematics or computer science can help data scientists have greater understanding of the algorithms they work with.
Because data science is a rapidly changing field, you’ll need to continue your learning by reading scholarly articles and staying on top of new developments. You’ll want to continue developing your skills and knowledge by working on projects to build out your portfolio. Once you do land your first job, you’ll have many more opportunities to develop yourself as a data scientist.
On the other side, there are also boot camps aimed at helping people with advanced degrees make a transition into data science. Since institutions of higher education have just started creating specialized programs in data science, many data scientists are people coming from other fields, such as physics, math, computer science, computational biology, economics, and political science to name a few. The data science boot camps that serve this group strive to provide the programming, big data, and advanced machine learning to supplement their domain and research knowledge.
Of course, looking at these types of boot camps doesn’t answer your question because it involves a path that includes both a college degree and boot camp training. It does however show the depth of knowledge that this sort of data scientist would have, compared to someone taking a purely boot camp path. For an employer, a boot camp graduate with no other education would be competing with someone with a rigorous research background. They would have an eye for bad models and poorly designed research. They’ve proven themselves as self-directed, self-learners. Certainly, they have their own hurdles to climb, such as proving that they can deliver business value and at a faster pace.
In general, in an either-or situation, you’re going to be better off with the college degree. You’re going to have a stronger and more well-rounded foundation. In particular, depending on your background, it may still be a good idea to look into a data science boot camp. Consider the following questions when deciding:
- How proficient are you already with one or more of the core competencies: statistics, programming, research methods, business domain knowledge, and communicating technical and non-technical audiences? If you have one or more areas covered, you’re going to be in a much stronger position.
- Can you learn on your own? There are tons of free and paid resources out there that can allow you to learn on your own, but it comes with its own challenges. A boot camp can be more efficient because you have a shorter feedback cycle in your learning with a pre-determined curriculum.
- If transitioning from another field, how related was your work to doing data science? If you’re coming from an unrelated field, you’re going to have a larger mountain to climb.
To determine whether attending a data science boot camp will be a good idea for you, check out this related Quora question: Are data science bootcamps worth it to get a data science job?
Originally published on Quora.