16
$\begingroup$

I have seen a lot of interview questions and answers for Software Engineering, Data Science and Machine Learning roles. I haven't found many interviews Questions for Operations Research. Can someone recommend a comprehensive way to prepare for interview questions and answers that are found for Operations Research

$\endgroup$
  • 2
    $\begingroup$ Are you interested in academia or in business? $\endgroup$ – Kevin Dalmeijer Jul 19 '19 at 8:18
  • 6
    $\begingroup$ My - probably very unhelpful - experience is that you best prepare for an OR job in becoming very good at OR. Ideally, you know a wide spectrum, maths, models, algorithms, software, how to implement, but also: practical needs, importance of data, application examples (and again ideally: you have worked on a practical application, e.g., in your thesis project). $\endgroup$ – Marco Lübbecke Jul 19 '19 at 10:57
  • 3
    $\begingroup$ I would add that PhDs and postdocs are likely to make you develop a few skills in extreme depth (combintorial optimization, integer programming, vehicle routing, location analysis, metaheuristics, etc), but versatility is not one of them. One thing that is largely appreciated is the ability for a candidate to think out of the box and to be comfortrable out of his/her comfort zone. A candidate must prepare to respond adequately to interviews that challenge this ability. I would add that this skill is of extreme importance in industrial interviews, but still important in academic ones $\endgroup$ – Claudio Contardo Jul 20 '19 at 16:10
16
$\begingroup$

First of all, ensure you can explain previous projects you have done very detailed. If you recently graduated this could also be your master thesis or papers you wrote as part of your PhD.

One of the best ways to test a candidate, in my opinion, is to do a case interview where we ask the candidate to solve an OR problem and go through the different steps.

An overall question could be "Let's say we want to build a model to schedule all our delivery trucks. What should we do?", I would then go through with the candidate in questions similar to these in broadly these three categories. (I won't provide answers to the questions, as I don't think it will help you in an interview to memorize answers to some specific questions. A good interviewer will keep probing to understand the depth of your knowledge.)

Problem definition skills

The ability to go from a loosely defined problem to a well-defined problem that can be solved using OR-methods. This also tests creativity and the ability to brainstorm potential problem components.

  • What could the objectives we would want to improve be?
  • What decisions should be taken by the model?
  • What constraints should we likely take into account?
  • What data should we try to gather?

Technical OR skills

The ability to build and code a model. This is where we can get really deep into the OR methods and technical details.

  • What method would you use to solve this problem?
  • How would you decide between using an exact or a heuristic method?
  • If we should formulate this as a mixed-integer programming model, what would your decision variables be? Can you write up a constraint that should be included?
  • If we solved this using a Tabu search, how would you create an initial solution? What would your neighborhood be?
  • What programming language would you use?

Communication skills

The ability to communicate findings and results to both technical and non-technical people.

  • Let's say we solved this problem and now want to present it to the responsible manager. How should we do that?
  • How can we visualize a solution?
  • How can we prove that our solutions are better than what they do today? What data would we need?

Hope this can help you. Good luck with your job search!

| improve this answer | |
$\endgroup$

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.