Operations Research Resolutions for 2012

It is the time of the year for resolutions.  Resolutions help define what we would like to be, but are damned hard things to follow through on.  As Paul Rubin points out, gyms will be overrun for a few weeks with “resolvers” (or, as my former personal trainer called them “resolutionists”) before the weight room empties again.  I’ve been as guilty as any in this regard:  it is no coincidence that my gym membership runs January 15-January 15 and that I renew it each year with the best intents, and the worst results.  Paul suggests an OR view of resolutions:

…the New Year’s resolution phenomenon offers a metaphor for O. R. practice. The “resolver” diets, exercises, stops smoking or whatever for a while because the “boss” (their conscience) is paying attention.  When the “boss” stops watching, the “resolver” makes excuses for why the new regime is too difficult, and reverts to previous behavior.  An O. R. solution to a business problem that is implemented top-down, without genuine commitment by the people who actually have to apply the solution (and change behaviors in doing so), is likely to end up a transient response leading to a return to the previous steady state.

So, in keeping with an operations research view of resolutions, I’ve been thinking about my resolutions with a particular focus on variables (what choices can I make), constraints (what are the limits on those choices) and objectives (what I am trying to accomplish).  It does no good to define objectives and go willy-nilly off in those directions without also defining the constraints that stop me from doing so.   But, of course, a creative re-definition or expansion of variables might let me find better solutions.

I have personal resolutions, and take inspiration from people around me who are able to transform themselves (yes, BB, I am talking about you!).  But I also have some professional resolutions.  So, here they are, and I hope they have been improved by an operations research view:

  1. Make time for research.  This might seem to be a funny resolution:  isn’t that a big part of what professors do?  Unfortunately, I have taken an administrative role, and there is a never-ending list of things to do.    Short term, it seems if I will be a better Associate Dean if I get on with the business of Associate Deaning, but long-term I know I will be a better Associate Dean if I keep active in research.  The decision model on time allocation has to be long term, not short term.
  2. Do what makes me happiest.  But where will the time come from?  I need to stop doing some things, and I have an idea of what those are.  I have been very fortunate in my career: I’ve been able to take part in the wide varieties of activities of a well-rounded academic career.  Not all of this gives me equal pleasure.  Some aspects (*cough* journals *cough*) keep me up at night and are not my comparative advantage.  So now is the time to stop doing some things so I can concentrate on what I like (and what I am comparatively good at).  While many of my decisions in my life can be made independently, time is a major linking constraint.
  3. Work harder at getting word out about operations research.  This has not been a great year for this blog with just 51 posts.  I don’t want to post just for the sake of posting, but I had lots of other thoughts that just never made it to typing.  Some appeared as tweets, but that is unsatisfying.  Tweets are ephemeral while blog entries continue to be useful long after their first appearance.  This has been a major part of my objective function, but I have been neglecting it.
  4. Truly embrace analytics and robustness.  While “business analytics” continues to be a hot term, I don’t think we as a field have truly internalized the effect of vast amounts of data in our operations research models.  There is still too much a divide between predictive analytics and prescriptive analytics.  Data miners don’t really think of how their predictions will be used, while operations researchers still limit themselves to aggregate point estimates of values that are best modeled as distributions over may, predictable single values.   Further, operations research models often create fragile solutions.  Any deviation from the assumptions of the models can result in terrible situations.  A flight-crew schedule is cheap to run until a snowstorm shuts an airport in Chicago and flights are cancelled country-wide due to cascading effects.  How can we as a field avoid this “curse of fragility”?  And how does this affect my own research?  Perhaps this direction will loosen some constraints I have seen  as I ponder the limits of my research agenda.
  5. Learn a new field.  While I have worked in a number of areas over the years, most of my recent work has been in sports scheduling.  I started in this are in the late 90s, and it seems time to find a new area.  New variables for my decision models!
OK, five resolutions seems enough.  And I am not sure I have really embraced an operations research approach:  I think more constraints are needed to help define what I can do, my objective is ill-defined, and even my set of variables is too self-limited.  But if I can make a dent in these five (along with the eight or so on my personal list) then I will certainly be able to declare 2012 to be a successful year!

Happy New Year, everyone, and I wish you all an optimal year.

This entry is part of the current INFORMS Blog Challenge.

6 thoughts on “Operations Research Resolutions for 2012”

  1. I come from the OR side and have lately been diving into the data mining side. Though I have a management position, I am grateful to my team showing me a new prism to look the world through. Your fourth resolution draws some sort of perspective of joining the two sides of the pictures, which I may have -inconsciously?- been hoping for.

    On another note, the combination of the Data Mining and OR goes beyond robustness: data mining helps one identify problem drivers, what an issue may really be coming from….in simpler words, it helps you better understand the problem, which OR may then address.

    Best wishes of success with your five resolutions

  2. Thanks for comment Semi Gabteni. Yes: I did not mean to conflate business analytics with robustness. They are related, but there is lots that is separate.

  3. Nice post! I particularly like resolutions 1 and 2 which I think for me come hand in hand. Despite being an early career research I already feel myself slipping away from 1 and 2, perhaps this year I’ll make a conscious effort to not 🙂

  4. pleasure reading the post. to me, resolution 2 seems to be on the top of my list. One of the best advice I ever heard about doing research is to love and be passionate about my research. I feel fornutate to be in the postion (still in graduate school) to focus on my research. Hope 2012 will be a productive year for me. Happy new year to all!

  5. My OR professor tells us every once in a while how his wife – who he claims is much smarter than he – does sport scheduling. He went on to tell us how many different combinations of round robin tournaments there were and how massive of a project it actually is.

    I just found your blog today as I’m in my last semester of MSOR at Kansas State University and still looking for a career.

Leave a Reply

Your email address will not be published. Required fields are marked *