Advice on how to get started for someone a bit later in life? [closed]

Question

I am 33 years old and graduated in 2006 from college with a degree in analytic philosophy. For a long time I thought that I was going to get a phd in phil and teach but that dream is long gone, and I am intent on having a career in data science/analysis.

I do have a history in mathematics; I finished calculus 3 in high school and finished a set theory course in college as well as advanced logic courses (symbolic, modal, predicate, etc), but I have never studied how to code, nor have I taken any statistics courses.

Are there any suggestions on how to best proceed? I of course want to be as attractive to employers as possible. I have no debt or family and am open and willing to take multiple years (and take on loans) studying for this profession.

Would getting a math degree (with an emphasis in statistics and taking courses like linear algebra) be the right (or at least, good) way to go? A local university offers a concentration in data science within the mathematics department, so I could do that as well. And in addition, taking courses in computer science to learn how to code?

Basically I'm looking to push a reset button (well not entirely since I have a background in mathematics, and I believe my background in philosophy will help too) on my career path and will do whatever it takes. I want a career with a future and with hopefully opportunity to choose where I work (city), and a career that fits with how my mind works (I believe I could thrive in this field).

Jobs that I have had since college will not help in any way towards future jobs in this field, so I will be fully leaning on future classes/preparation.

Thanks in advance and for reading this partial ramble! I'm motivated but don't know exactly what I should do. I realize that there are many paths people take to get into this profession but would love advice for someone like me in my situation.

Answer 1

I for myself wouldn't advise you to not take that route. I do agree with most of the points that were mentioned before but my conclusion is a slightly different one:

Trying to get into data science without any preceding field knowledge deserves respect and it is not impossible. Also, there is still a lack of qualified data scientist and I would predict that the demand is growing.

But you should consider the following points:

• Data scientist is probably the job in the statistics/IT area that requires the highest level of education. Ask yourself if this is really what your aiming at.

• If you choose to go on that route be aware that it will be extremely hard. Data science is so multifaceted and interdisciplinary that there is a lot to learn. Even if you'd have a couple of hours to study each day I'm afraid it would take years. Being able to work as a data scientist needs far more than an academical math knowledge.

• You will have to compete with PhDs in their mid-twenties who already have years of experience in publishing and conducting machine-learning/math/programming projects. Employers will probably favor them over you.

If you don't care that it will be painful and want to start learning, I'd do the following:

• Get basic knowledge of statistics, programming, and IT-infrastructure. You should be familiar with t-tests, ANOVAs, regressions and cluster-/classification techniques. You should start learning a statistical programming language. For starters I'd recommend R or Python. Basic knowledge of another higher programming language is useful. Knowing basic programming principles such as object-oriented programming doesn't hurt either. I'd recommend to take a look at Java since a lot of big data tools such as Hive, Hadoop or Spark are Java-based frameworks. Look at these frameworks, too. Understand what they are used for. At last you should have at least basic knowledge of databases and know the difference between a relational and a non-relational database.

• Don't spend money on courses. The internet is full of free sources. Just take into account who offers them. There are a lot of lecturers and professors from very good universities who provide online courses or share their lectures.

• Read books. Again: Don't spend money. Search the internet for free high-quality materials (see here for example).

• Check out community sites such as Kaggle and join competitions. This is probably as close as you will get to real-job tasks.

• Continue doing that for a couple of month and ask yourself again if this is really where you want to go. If yes, continue.

But realistically, here is what I would try to do:

• Don't aim at becoming a data scientist right away.

• Consider to start in a statistics-related job (e.g. controller, business analyst, marketing manager, game designer etc.) and work your way up. I strongly believe that such a strategy is more likely to be successful.

Because no matter what you do before starting in the industry, you will have to learn on the job.

Answer 2

I wouldn't advice you to go that route. You're saying you'd like to study for a few years and then start as a beginner (and what other than a beginner can you start at? or a graduate programme). But bear in mind by the time you're 37, there'll be 22 year olds who'll be starting as a beginner just like you, with perhaps comparable skills (though perhaps not your educational background). Why would an employer take on someone fifteen years older?

Why would you like also to be as attractive to employers as possible? Do you think that leads to anything other than employers wanting to use you? And what do you think happens after you're used--unless you'll make damn sure it will not? You probably want to be attractive to either get approval, remuneration, or advance in your career, but these three don't go hand in hand. It's difficult to get them in equal measure, e.g., often to advance in your career you have to take extra risks, like resigning, or conversely, to get accepted for more jobs lower your expectations. I don't see why you'd like to be attractive to employers and assume that that'll take care of your success, it won't, or it might not. Why not, rather, think of working for yourself? Find a career in which you can achieve independence. Specially as you get older.

Finally, I think it's tricky to try and build a career in a field like data science that's new, 'moving ground' so to speak, and not quite well defined (for some), essentially by way of courses and university studies. If you don't have the computer industry background to know the [future] value of what you're sold, how are you gonna choose? There're plenty of places ready to sell you a course, and many will use vivid depictions of future success or 'professionalism' in order to attract you (because, for these academies and schools, you're the customer, and of course they want / need customers). How do you know you're not going to be spending time and money taking lessons in something that'll be perfectly useless? And why take courses? Since many of these topics, disciplines, and programming languages, are free to learn online? Or do you assume that studying at a certain place means that that place will take care of your career, or necessarily give you a good foundation? Why not, rather, find out how present-day data scientists got to where they are? I think, if you interview some, you'll probably find no two paths were alike. And you can't emulate any of them (because what worked when they got started is not necessarily what will work for you tomorrow).

To me it sounds like you'd be more adept to teaching than at data science. This is because you sound like you like to learn, like to teach (to advance yourself and others, I'd like to think), and you say you have time to do it. I'd try and get into teaching through some way other than that PhD in Philosophy you mentioned. Maybe start your own academy, alone or with a partner, and organize yourself to teach others what you already know, or what you're capable of learning, or teach them how to learn, or at least how to strive (motivation). That would sound to me possibly like a better investment.