I am a fourth year Software Engineering Student, i always dreamed of becoming a mathematician.
Doing math is my passion, my happiness, my everything.. but for some reasons i couldn't do it.
I am now a software engineering student with a high GPA( i reached 87/100) but i am not happy.
Until that day when i was attending a conference did by Facebook Developer Circle in Lebanon(I am a
member) the subject was the Future Jobs of Developers. They talked about Data Science and the
importance of mathematics in this field. I realized then that my dream can be achieved.
I started with https://www.dataquest.io and EDX.
I am asking for an advice.
Thank you in advance.
While it is possible to do data science with little math and stats as I have done for more than 20 years despite my strong statistical/math background, most employers will look for someone who knows classical machine learning techniques pretty well. Being able to write professional code in Python is a big plus, and some statisticians lack this skill. So is a deep understanding of how various database architectures work, and having worked on real, complex data. This should be an advantage for you. A data science role that would fit nicely for a software engineer, is data engineer or architect. Also many data scientists use libraries to perform stat/math procedures (gradient algorithm, regression.) However, many employers -- during a job interview -- still expect candidates to know the details about these procedures. While some are unimportant (eigenvalue theory), some are (how to interpret results, differentiating between correlation and causation, when these techniques fail, cross-validation, model-fitting, and feature selection done right.) In some "data science" roles (BI analysts), you might only work with Excel and SQL. Some employers will provide any training that you lack, and in some cases, they like someone new to the field, as these employees do not have to un-learn some bad practices acquired during college years. Physicists are routinely employed in data science roles, for instance in Fintech, so the field is definitely open to people with various backgrounds.
Thank you Mr Vincent for you answer. So after 20 years of experience what is your advice as a software engineer(i know python and many programing language).
Is it possible to get a job as a data Scientist ?
Vincent summed it up nicely.
Mariam, one other thing you could do as an undergraduate to help demonstrate your mathematical strengths is to participate in a mathematical competition. In the US and Canada we have the Putnam competition, and elsewhere in the world there is the IMC (http://imc-math.org.uk/).
Thank you Mr Justin, i really appreciate your advice and i always look for competitions but unluckily in Lebanon we don't appreciate the field of mathematics. There isn't community or programs to encourage student, i study online courses from outside Lebanon.
Understood. An alternative to demonstrate your data science chops is to sign up at Kaggle (https://www.kaggle.com/) and compete in one of their many data science competitions. Consider joining an existing team, too: it's good experience, exposure, and networking even if your team does not win.
If you manage to win a competition or two, your next job interviewer is not going to care what your major was or where you studied because you've demonstrated the ability to solve data science problems in a very public way.
I concur with all of Vincent's advice. As a statistician in the ci/data analyst in the civil service, my trajectory is a bit different. I would be less concerned with applied mathematics and put more emphasis on probabilistic theory and statistics, particularly regression analysis, time series models and forecasting, matrix algebra, etc. Understanding the practices and theories behind these can be extraordinarily useful, industry-wide. Your software engineering background can be very useful as you may find yourself developing in some capacity. I hope this useful is useful. Good luck!
Thank you Mr Karriem for your reply it's very helpful :)
Being a data scientist meaning that you will be able to translate number to be a problem/solution and theory, while the abilities needed for this is not just mathematics. In the complex datasets, there're a lot (more) to understand the link between factors and variables rather than just (only) know how to run regression or modeling or ANOVA or pearson's correlation etc. So i think another most important thing of being a data scientist is being able to interpret the results and translate those into theory(ies) that really can explain something better.
PS: most of people will see numbers as numbers, while data scientist always see numbers as a problem to solve, how and why
So as a result i see being a data scientist requires this mathematical mindset which is what i find helpful.
Thank you for your reply Mrs Satdichanh:)