What's Your StORy? Paul Rubin

Name: Paul Rubin

Affiliation: Michigan State University (retired)

Title: Professor Emeritus

INFORMS member since the early 1980s (first as a member of TIMS)

Happenstance. I was working on a PhD in “pure” mathematics (convex analysis) under a faculty member who at one time had held a joint appointment between the Math department and the Management department (where Management Science was housed). Management lost a faculty member and needed someone to cover some courses. They asked my advisor, but he declined and offered me up instead. So for a year I taught graduate management science classes (including teaching other doctoral students, which was a bit of a trip). It was my first exposure to management science and O.R., and I liked the applied aspect (but did not yet see it as a career). 

After that year, Management hired someone and I returned to Math to finish my degree, with plans to work in industry. About when I was hitting the job market, the new Management professor left and they were again hiring. A senior faculty member there talked me into applying, and after their first choice turned them down and their second choice didn’t pan out, they were stuck with me.

There are significantly more women since I started, although we are still well short of 50 percent. There are more members of minority groups as well, but we have even more room for progress there. Computing has come a long way. (Bear in mind that I started in the punch card era.) Software is more accessible, and we have pushed a lot of introductory O.R. and analytics content into “functional area” courses such as supply chain management. That increases its accessibility, but maybe reduces the demand for members of the wizard class.

First, learn to program. I’ve met a few people who assumed that they would be able to dump the programming chores on someone else (professional programmers in industry, student programmers in academe). That’s not always true, especially when you are in the early investigative stages of research, rather than pumping out a finished product. 

Second, learn to compute, which is not the same thing. You need to understand the limitations of floating point arithmetic, how memory structures can make your algorithm faster (or slower), what creates numerical instability, and other things that can cause your brilliant algorithm to spit up horribly wrong answers (or melt the CPU).

I have not been matched with anyone, which is perhaps a good thing since I’m retired. That said, there are a number of reasons for young members to seek out a mentor. First, you might not find a mentor in your home department, either because the potential mentors are too busy, personally incompatible, or just not interested. Second, it can be useful sometimes to get an external opinion from someone who does not have skin in the game. Third, if your INFORMS mentor is inclined to introduce you to other people at the meetings, it’s a chance to build your network of professional contacts.

First and foremost, it’s a chance to help people who are themselves helping people. So there is the potential for warm fuzzy feelings, and maybe accruing some positive karma (which some of us desperately need). As someone pointed out (I forget now who), the value of our donated expertise probably exceeds the value of any monetary donations we might be inclined to make. 

Second, and I think this is a particular benefit for student volunteers, you get to see what things look like in the real world. Textbooks (and most classrooms, I think) create little bubble worlds in which all the data you need for an exercise is handed to you (and only the data you need), it is all correct, your goals for using it are clear, and the correct analytic method is whatever the current chapter is covering. In the real world, and particularly with small nonprofits, there may be little or no data yet collected, it may require collation and extensive cleaning, and what you then do with it is largely up in the air. So even the most ordinary projects can be learning experiences as well as opportunities to “give back.”

The meetings are my favorite benefit, both for the networking opportunities and for the chance to learn new things.
 

I’m going to go with the coffee breaks and the exhibit area, mainly because I get a lot of networking and ad hoc contacts done there. The social events are a lot of fun and also provide opportunities for networking, but it’s harder to network when everyone is crammed into one area (and food is being crammed into your mouth). 

The INFORMS Journal on Computing is my favorite, because more of the articles are relevant to my research interests (applications of integer programming, without a focus in any one application area) than what I find in the other journals. My second choice would be Interfaces (now the INFORMS Journal on Applied Analytics), for the application articles. 

Working out at the gym, Taekwondo (smacking things can relieve stress, so long as they don’t smack back), and vegging out in front of the TV (but most definitely not watching political news).
 

The giant tortoise. It’s resistant to be steered or led around, and if things get gnarly it can just retreat into its shell until the storm blows over.
 

First off, the premise is faulty. I’m a geezer; nobody expects my socks to match. Second, I pretty much only buy white sweat socks these days. Third, you haven’t told me if the various socks feel alike. However, I suppose the answer you are looking for is 19/44 * 18/43 + 25/44 * 24/43 = 0.4979 (approximately).