4
$\begingroup$

I'm an undergraduate majoring in economics, and I am deciding whether or not to take real analysis. My plan is eventually to get a Master of Science in data analytics, similar to this program:http://analytics.ncsu.edu/?page_id=1799. Given my course load, I may be able to make it fit, but I may not. This will be the last time I am able to take it. The course description reads: "Studies continuity, convergence of sequences and series, differentiability and integrability. Introduces appropriate topological concepts. How helpful will this class be in pursuing a master's in data analytics?

$\endgroup$
  • 7
    $\begingroup$ Depending on what your definition of "data mining" is, the answer can vary from "extremely useful" to "almost useless". There are numerous questions with thoughtful answers on related matters here on this site. If you haven't had the chance to peruse them, I'd highly recommend that and also editing this answer to clarify your ultimate objective(s). For example, if you have any interest in pursuing graduate education in economics or an allied field, I would strongly recommend taking such a course independent of any interest in "data mining". $\endgroup$ – cardinal Jan 4 '14 at 19:04
  • $\begingroup$ If your objective is to eventually study data analytics, good programming skills are vital. If you don't have them, focus getting them instead. Data mining is a huge and diverse field, and as Cardinal said, we can't answer that unless you specify what route you want to take. $\endgroup$ – rocinante Jan 5 '14 at 16:05
3
$\begingroup$

"The curriculum is a carefully calibrated mix of applied mathematics, statistics, computer science, and business disciplines."

Based on this taken from the link you provided, and my personal experience with data mining and statistical machine learning, I would say it would not be worth your efforts. It seems like the program you are interested in is much more applied and would not go into the mathematical rigor that would require a prerequisite knowledge of real analysis. Could it be useful for understanding the fundamentals of calculus, sequences, and series? Sure, but you could learn this without necessarily needing to learn how to rigorously prove everything. I would spend more time with programming and linear algebra to prepare you for data mining. Unless you want to go into acedemic research in data mining at the Phd level and beyond, I would not take real analysis.

$\endgroup$
1
$\begingroup$

Based on the description of the program in the link, I would say that it is not the most important thing.

If your plan is to stop at the masters level, then real analysis may not be important. However, if you are considering a future PhD that may officially fall into a statistics or mathematics department, then it might well be worth your time to take it. Departments like these do require real analysis--I have been associated with several universities (U.S.), I do not know any programs that do not require it. The masters level focuses more on practical skills than theory, and real analysis is more important for theoretical work, deriving methods, and bracketing results. Just to be clear, I'm saying it may be important for coursework, not necessarily for actual work.

I agree with other posters that programming skill may be more practically useful specifically for data mining. Though I am a big supporter of sound theoretical understanding.

$\endgroup$
1
$\begingroup$

If you want to get a master degree and eventually go to industry, I don't think you need take real analysis course. It is kind of too theoretical and far away from practical data mining techniques. On the contrary, multivariable calculus, linear algebra, statistics and optimization knowledge of undergraduate level is more helpful to appreciate almost all algorithms.

However, if someone wants to do Phd in related area, e.g., machine learning or learning theory, it is totally recommended, even necessary.

$\endgroup$
  • $\begingroup$ Real analysis provides a solid theoretical background, as Hua mentions. It is not necessary, but it will often take you far regarding other problems. Although you may not use the techniques from the course directly, it may help you to conceptualize ideas and think outside the box. $\endgroup$ – Gumeo Nov 2 '15 at 15:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.