I began my graduate study in the 2nd semester of the year 2006. Back then, I was deciding between two OR tracks at UP Diliman, one offered by the Math Dept. and the other by the Industrial Engineering Dept. I emailed the former and I was told that the focus is more on the theory and if I intend to do practice, then I should go with the latter. Since it was clear to me that I intend to be an OR practitioner, I took the advice and headed to apply at the College of Engineering.

Fast forward to today with less than one year before graduation. Although I have already taken 30 units of graduate (and refresher) courses, I still feel that the MS IE/OR program doesn’t sufficiently equip its students with the skillset for successful OR practice. It does cover the basics but some changes are necessary. I got this sentiment when I read a career article entitled “What Industry Wants from OR Grads.” Some of the skills deemed to be important include

  • database (e.g. MySQL, Oracle) and statistics (e.g. SAS, SPSS) software; manipulation and analysis of data are important because only raw data is available initially, and not information
  • programming (e.g. C++, Java)
  • modeling (ability to prototype and interpret the outputs of various OR models)

The first two skills are not part of the standard MS IE/OR program, though it’s possible to take elective courses to develop these skills. Modeling, on the other hand, is a critical OR skill but it seems to be not getting enough attention. Clearly, there’s a gap between the skills developed in the academe and the ones needed in the industry. Why is there a gap? Based from my own experience, here are some of the problems with the current MS IE/OR program:

  1. lack of case studies and interesting problems
    Case studies provide motivation and a clearer picture of how OR practice is actually conducted. They are important because lectures can tend to get too theoretical, especially with unrealistic assumptions. Sometimes, I tend to ask “Is this tool actually being used in practice?” Also, the models discussed were limited to textbook problems only. More engaging examples can be found on the Internet, such as solving Su Doku with an integer programming model.
  2. too much focus on the algorithms
    Although it’s important to have an understanding of the algorithms for output interpretation and devising heuristics, I believe model building is more important for OR practice. As I mentioned in a previous post, I was amazed that millions of variables and constrants are not uncommon in industry applications, while the largest I have studied consists only of hundreds. I’d like to see these “monster” models discussed in class and see how they can choke the solver software (maybe in our Advanced OR class).
  3. insufficient human resources
    All of my graduate subjects thus far were conducted weekdays at night or weekends because my professors either have consulting or daytime jobs. I’m very much grateful to them because they lent their time to teach us what they know. But I do think that a full-time professor may be more effective in teaching because he/she can spend more time in research and do additional lectures beyond the scope of the syllabus.

Final thoughts

Even with these problems, I believe that the MS IE/OR graduate is equipped with enough competence to learn the necessary skillls on the job. Still, improvements are necessary to bring the program to world-class standards.