Do I Have to Know Math to Enter the Data Science World?
My personal answer to this FAQ.
It’s been over a year since I started studying data science. It all began for the sake of curiosity, but then I gained a pleasure that made me study for a career leap (not accomplished yet).
As I study and practice almost every day, the love I feel for this field gets me to share the knowledge I get (and, of course, I do it even for the pleasure of sharing).
Sharing information on Linkedin gets me to receive connections requests, mostly from people who want to know more about data science, because they are curious and want to understand if this field can fit them. The typical question people ask me is: "do I have to know math to get into data science?" Or: "How much math do I have to know?"
I know this question is very common, and I want to give my personal answer to people who are starting to study.
1. You don’t need to know math to learn to program
A big surprise in the beginning: if you want to study data science you need first to learn to program.
Take a course and just start programming. It doesn’t matter what course and what programming language (I advise Python, as it is mostly used even for non-DS things, and you may just like Python and not DS).
But the thing is: you do not need to know math to learn to program. Programming is just…programming! It’s not "exercising on math". Of course, if you have the habit of solving logical problems you will be at an advantage; but, you know, the ability to solve problems is a general ability and you learn it just by doing.
My specific advice is to start from the beginning. Learn to program from 0. Learn functions, lists, tuples, ad so on. Cover the basics for some weeks and when you are a little satisfied, go on.
Of course, if you are like me you may have a little difficulty; in fact, when I approach a new field and discover I’m loving it, I want to know more and more every day; I have a hunger for knowledge and I have to fill is as soon as possible. In this case, the game to play is very simple: learn what you need while solving real problems of your own.
The beauty of DS (and, more generally of programming) is that you can create your own problems and solve them, gaining experience.
So: start programming from 0, and then "create your own problems to solve". Do you need to do some calculations and report the results, appending them to an excel file? Good: learn how to do it! It is a very good exercise to gain experience; trust me.
2. Start analyzing some data
After some weeks spent learning to program, start analyzing some data. You have a bunch of possibilities to do it: get some courses, go on Kaggle, and so on.
Even here, it is not important the way you decide to start: just start! You can change the course if you do not like it or if you find it too complicated (or too easy) for you!
The question here is: you need to know some math. Yes, my friend, I had to tell you: to analyze data you need to know math; don’t believe anyone who tells you otherwise. But there is good news: if you don’t know math, you can learn it while you need it!
At this stage, you will need the basics of statistics: mean, median, normal distribution, and things like that. If you do not know them, learn them for yourself when you need them. There is a bunch of information on the internet (and thousands of math courses): learn what you need when you need it, and practice doing projects, so that you can master the theory with practice.
If you feel overwhelmed, do not be scared: you don’t need to know everything.
I’ll tell you a secret: no one knows everything!
Repeat it with me: no one knows everything. This must be your mantra, otherwise you’ll get lost in the field.
There are a lot of people out there who know a lot of math and statistics (a lot of Data Scientists come from a mathematics/statistics background); the truth is: a lot of these people come from the research world (often, they have a Ph.D. degree), but the "everyday reality" is simple and you do not need to know the math a statistician knows to be in Data Science!
So, start simple and go on practicing every day, studying what you need when you need it, and in some months you will see the results; trust me.
Yes, I have an Engineering background, but I have never attended a statistics course during my Engineering studies: I’ve learned it on my own while studying Data Science, and I’ve studied the topics when I needed them. It works, trust me; and you won’t have the feeling you have to know everything, which may freeze you and gets you to leave your studies.
3. Machine Learning
I want to be clear from the beginning: to practice Machine Learning you have to know a lot of math. Machine Learning "does things" and "behind those things" there is a lot of math.
You can not apply algorithms to datasets and pray for the results to be ok; this won’t work. Remember: garbage in, garbage out (your second mantra!).
Machine Learning algorithms can not be a black box to you: you have to understand what you are doing! But there is good news: you can learn the needed math while learning Machine Learning.
A typical example I do is Gradient Descent. If you have studied calculus, when you hear the word "gradient" you know we are going to talk about derivatives, tangent lines, and maximum/minimum points. But that’s it! If you don’t know what a derivative is, you don’t need to solve 876 exercises on derivatives (as I deed in High School and at the University): just understand the topics you need and go on. Then, practice, practice, and…practice mastering algorithms.
Conclusions
If you want to get into Data Science and you are asking your self "do I need to know math to get into this field?" my answer is: no, but yes.
No, you don’t need to know yet all the math you need.
But yes, at a certain point you need to study math you need to understand the topics related to Data Science.
Start by learning to program and then go on, every day, learning what you need (with respect to programming and math). Don’t be afraid of what you don’t know: no one knows everything. Just start and see how it goes and…enjoy your learning path!
I know you can feel overwhelmed: start with a Machine Learning course, get an Algebra course, return to Machine Learning, take a Calculus course, etc…but remember: the learning process is like building a house: brick after brick, starting from foundations, you will arrive at the roof. Day by day, without haste and without respite. There is no other way.
Let’s connect together!
LINKEDIN (send me a connection request)
If you want, you can subscribe to my mailing list so you can stay always updated!
Consider becoming a member: you could support me and other writers like me with no additional fee. Click here to become a member.
Share This Article
Towards Data Science is a community publication. Submit your insights to reach our global audience and earn through the TDS Author Payment Program.
Write for TDS
