0
$\begingroup$

I am a new graduate student in economics, and my advisor has recently focused on applying machine learning techniques to economics. It seems natural for me to take up this area, but I heard the rumors that this field is highly immature. It is believed by some others that the researches existed are unsystematic and disorganized, and most researchers are of the type "Try this. If it does not work out, try that." They also claimed that working on this area does not help if one wishes to step foreward in academy.

I am not sure what to do, and I really need some expert advice. Can anyone confirm or disprove the argument above? Also, are there any summaries that worth reading, or any researchers in this area that I may follow up? Thanks in advance!

Improve this question
$\endgroup$
3
  • 1
    $\begingroup$ @Giskard Sorry for the ambiguity. For data science I actually mean machine learning and deep learning. $\endgroup$
    – zyy
    Apr 14 at 11:52
  • $\begingroup$ It probably won't answer your question but still, if you are a new student, it would be useful to go through this document. annualreviews.org/doi/pdf/10.1146/… $\endgroup$
    – mark leeds
    Apr 14 at 16:17
  • $\begingroup$ also, check out the references. the paper by Breiman (2001b) is extremely well known. $\endgroup$
    – mark leeds
    Apr 14 at 20:01

1 Answer 1

Reset to default
1
$\begingroup$

The thing to remember is that a lot of econometrics is concerned with establishing causality between a variable of interest and the outcome ex-post. In this framework, establishing unbiased estimates of key parameters is the goal, and predictive power is less important.

Much of machine learning is concerned with minimizing error in prediction, with the bias of estimators being of less concern, so there is a little bit of a mismatch in terms of the historical goals of econometrics and the machine learning literature. That said, I am sure you can come up with important applications if you choose to do research in this area. As a young researcher, an "immature" field is beneficial to you since it will be easier for you to contribute. The Athey and Imbens paper shared above is a very good starting place.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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