Let's Start With the Thesis

As promised, I will tell you about what's been happening in my life over the past two or so years in the next few posts. The Master's thesis is almost certainly the most significant, but least useful of these events.

Let's go back to 1980...

I am in grad school at the University of Illinois, Urbana-Champaign, working on an MS EE degree. My proposed thesis is on mixed-mode digital/analog simulation and all of the programming is done. All I have to do is to write the thesis. However, in my personal life, all is not well. I am working the equivalent of a forty hour a week job while doing another twenty to thirty hours a week of work on my thesis topic. As a result, my first marriage is breaking up and I want to try to make a fresh start with my wife without the additional hours of my additional work weighing me down. In the industry, the (as it was then called) Computer-Aided Engineering market is booming and people are being hired just out of school with (what seemed to me at the time) princely salaries. I am offered a job with Tektronix in Portland, Oregon, and having fallen in love with the area, I move out there (and still my here) with a promise on my lips to my thesis adviser (Professor Jacob Abraham, now of UT Austin, and still a prince among men) that, of course, I would finish my thesis. A year later, my divorce is final and I have not finished my thesis. This state was to remain unchanged for the next twenty-seven years.

Fast-forward to 2004...

I've made it through the dot-com bust relatively unscathed and the industry is picking up a bit again. I'm getting to the point where I'm thinking of teaching again (I taught EE as an adjunct instructor back in 1983, when folks still thought I was going to be finishing my thesis, and I enjoyed it). However, checking around, I find that the entry level teaching degree, even at a community college is an MS. I start thinking about going back to school. In addition, I am also getting to the point in my life where I'm trying to put behind me some of the regrets I've had in the past. One of the largest was my inability to finish my thesis. To make a long story short, I enroll in the Oregon Graduate Institute of Science and Technology) in the "Adaptive Systems" Program in the CS department and finish my required classes in about a year and a half. In a fit of insanity, I ask the head of the Adaptive Systems Lab (also, in the end, a prince among men) to be my thesis adviser.

It is now sometime in 2005-2006...

Time has not been kind to OGI. Having over-expanded in the dot-com boom, it is now in dire financial straits and is forced into a shotgun wedding with Oregon Health and Sciences University (our state's medical school). From this point on, all research needs to have a biological tie-in. This is not great for me, as I really have little interest in biology. My adviser, having seen the writing on the wall a few months before me, has started working with a team at OHSU investigating the signal processing capabilities in the neurons of the electro-lateral line of the Mormyrid fish. More precisely, he has noticed that perhaps, when one creates the equations that describe the statistical behavior of ensembles of neurons whose learning rules are of a certain form, one might be able to find a closed form solution for the moments of the equilibrium probability distribution of the neural weights. This points an arrow back to some of his earlier work showing that approximate solutions for the equilibrium distribution can be wildly wrong and that alternative formulations such as this one (and more importantly, his prior use of perturbation equations) should be pursued (of course, with the use of copious NSF grants). The main thesis point I prove over a period of about three months in the winter of 2006-2007. The remainder of the year is spent attempting to understand (a) biological neural models, (b) statistics, and (c) why anyone would give a rats ass about this (other than for the copious NSF grants). I eventually figure out (a) and (b), but remain perpetually unenlightened about (c), finally producing a thesis entitled "Exact Moment Dynamics for Spike-Timing-Dependent Plasticity" which is critiqued endlessly by my adviser over the rest of 2007 and beginning of 2008.

By this point, my current wife is getting increasingly cranky about the lack of progress on the whole "graduation" part of this process. I have spent approximately $30,000 on this endeavor so far and must continue to enroll in at least one credit of "Thesis Research" (read ~$600) for each additional term I remain in school. The wife is understandably getting nervous that I will be working on this forever, or until the School of Engineering is shut down (Did I also mention that with every passing term, the headcount and facilities of the School of Engineering is shrinking?) and I am left at the mercy of remaining faculty who understand Biology, but not statistical mechanics. She gives me an ultimatum - finish or quit. Luckily, with a final heartfelt talk with my adviser (Did I mention that he is a prince among men?), we agreed that my thesis *with final minor edits" was fine and we set up a defense for August.

To make a long story short, I defended my thesis in August and got most of the readers' signatures I needed to pass my defense, including that of one professor who left for Massachusetts the day after I presented my defense (I did tell you about the shrinking head count, no?). However, one reader wanted a minor edit so I was not able to get his signature until September 2008 (which required that I pay another $600 for another unit of "Thesis Research") and, in the meantime, the school had changed the format of the form on which one got the signatures for the defense. More worryingly, I needed the signatures of one professor not associated with OGI/OHSU to graduate. One of the professors was associated with Portland State University and OHSU when he agreed to read my thesis. However, a week before I finished my edit/got the signature, this professor was laid off from his PSU appointment and, as such, was only associated with OHSU and no longer qualified as a non-OHSU reader.

This leads to the final, knuckle-biting part of the story, where I submit the thesis with the old defense form, and with the signature of the now non-non-OHSU reader, having been assured by my adviser that, of course, this was alright. It is getting to November and I have still not heard whether or not I had met the qualifications for graduation. I go to the graduate studies office and ask the woman there if there was anything else I needed to do or if anything was missing. She let me know that this was her first time she had done this (Did I mention the loss of head count in the School of Engineering?) and that she did not even plan to look at anything until after the Fall term had ended. Of course, by mid-February, nothing had happened, I still had received no notification and, understandably worried, I inquired again. She said that it shouldn't be more than a couple more weeks. Finally, at the end of March, a "Graduated in December 2008" appeared in my unofficial transcript. They still didn't send me my diploma until June.

So that's how someone who had no interest in Biology or Statistics ends up with an MS Comp Sci focused on Biology and Statistics (two fields in which he still has little interest). Having been a software development manager for the past ten years or so (mainly in the areas of computer security and system management), I was under no illusion that this would help to advance my current career whatsoever. I am also no longer enamored with the notion of teaching. I have met enough academics in this troubled period to understand how insecure their positions really are, how petty is the institutional infighting, how draining is the constant begging for funds and, in most cases, how thankless their work is. I have no desire to attempt to make their lives more difficult by adding to the pool of potential instructors. In the end, I met a few more people who I liked a lot and I closed out some unfinished business that had been hanging around for twenty-some odd years. If there's any grand insight to be gained from this experience I guess this is it: "If you hang about the computer industry for thirty odd years, you get to be good enough in things you're not that interested in to be able to do almost anything."

Not an edifying lesson, but oddly appropriate...