MPhil leading to a... drumroll... PhD

Well, it's official: I'm a research student.

The teacher of one of the classes that I was attending in the MSc course asked the group who was going to or thinking about doing a PhD in the University of Manchester after finishing. I think it's not needed to write that, well, I was totally there (hmmm, if it's not needed, why did I anyway? Think about it).

The teacher's name, by the way, is Dr. Barry Lennox.

His logic was very convincing: what's your reason to be doing a MSc, other than it being the precesor of a PhD? If it's to get knowledge before hand, that's what you do in your first year of the PhD anyway: actually, the program that you can get into if you don't have a masters degree is "MPhil leading to a PhD" which will grant you a MPhil in the first year.

Brief parenthesis: there's a big difference between a MSc and a MPhil. A MSc is a taught program, meaning going to classes and doing a small dissertation project in the last three months of the 15-month period. A MPhil is a research program, meaning reading, reading, reading, and doing a small thesis about the findings at the end of the one- to two-year period. Both give a masters degree, but a MSc degree is a little more prestigious than a MPhil in the U.K, mainly because of the fame that it has brought on to itself of being "the degree that is given out when someone doesn't complete a PhD". Although, in my perspective, being that a MPhil is research-based, in Mexico, where it doesn't have that fame, it should be the other way around.

And that leads to Dr. Lennox's second point: if some reason you don't find any funding for the last two years (which, considering the 300% increase in tuition fees for foreigners, is viable), the University will still give you a masters degree to go back and boast about. Taking care, at the same time, of the fact that you'll have to pay for just three years of tuition fees, not four (£11,000 less to pay is always good).

To finish off his argument he mentioned that if you're seeking funding, you need to apply during this year to have the whole process finish by the time you finish the MSc (around the end of next year). And if you're applying for help to fund a PhD, the scholarship organizations will ask for a summary of the project that you're going to be developing; if you have this already, why not start working on it now?

It was hard to ignore the opportunity, but one thing was bugging me: I didn't have a project to work on. That was the point of doing the MSc first: get an idea of what can be done. I went to talk to Dr. Lennox anyway, and he first asked of my qualifications: I mentioned my bachelor's degree GPA (94/100) and that I have my own company, Makko Solutions. Apparently that was enough for him: "You're in. What projects do you have in mind?" I was completely stumped. I explained the whole situation and he replied that I should think of some ideas in the next couple of days, of what interests me, what area is the one that I would like to get into, and to read what the staff have dwelled about to also get a feel of who would I like as a supervisor; I wasn't even a research student, and I already had an assignment.

I was actually quite impressed with the amount of ideas that came about. I never realized that I have had stored so much stuff in my little nugget. I first read of what the staff have written about, which was glaring: very diverse stuff, really. But I settled on two: Dr. Heath and Dr. Lennox. They have written about a subject that I've always heard about and that sounds interesting: Model Predictive Control, MPC, and justs as it sounds, predicting to satisfy every state of the controlled process (to read more: go here). It also implies controlling non-linear models (which in classical control is VERY hard) with more than one input/output (also a pain).

The ideas that I most liked where the following:

1. Using MPC as an optimizer of the process of coming up with a solution by genetic algorithms.
2. Global-Positioning-System-based guidance system as a control problem solved by MPC.
3. Quality control by MPC.
4. Flying Saucer project.
5. Using MPC for fault monitoring in the Intelligent Guitar Effect Box project.

All of these are quite hard to explain in layman's terms, because of the specialization that they involve, but I'll try:

The first one. Genetic algorithms is an artificial intelligence concept, which, of what I've read, it's the standard of comparison for other problem-based solution-finding algorithms. The problem is it's speed, because it relies on evolving over all the possible scenarios (no matter how improbable they are), so I'm proposing using MPC to optimize it and speed it up by trying to predict the way it will evolve the quickest. The interesting point of this is that usually AI is used as a tool for control, but this proposition reverses that, using control as a tool for AI.

The second one. Putting the reference point of the system to be the desired position of a vehicle and the observed variable being it's current position (all this GPS driven), use MPC to not only solve the positioning problem, but also try and predict problems and obstacles ahead.

The third one. If you're from the industry, quality control must be day-in and day-out concept for you. I would like to see if MPC fits in there and what advantages can it bring out.

The fourth one. I always thought that flying saucers were possible to make: by a fan controlled by a motor and counterweights in it's edge, not by the center. MPC will be used to handle height control and direction.

The fifth one. This was just to try and include my hobby in the whole thing: imagine a guitar effect's box that you teach your song to, and the box will learn WHEN, in the song, to put WHAT effect. Now, when you go on stage, just plug it in and you don't have to come back to step on pedals to apply the next effect. Obviously, because we always play a song differently everytime, the box needs to predict and compensate for any error in tuning and/or timing whilst playing live.

I met with Dr. Lennox with the ideas in hand, and was very surprised that he was very interested in the fifth one... yeah: the guitar one. He mentioned that he had started working on a game which would teach a person how to play a certain song, like the dancing games you see in every arcade, but with any type of guitar. That actually fitted well with my project, but he instated that it's very important to keep an open mind on where the project may lead. It's very possible that this project, or it's findings, may be applicable to other areas that aren't sonic analysis, and, thus, don't get too attached to it because it may change over time. Almost as in queue, here comes a bank, yes, a bank that has grown interest in the project and it's feasibility of applying it to credit account owners and/or the accept/reject process of getting one.

It's now being about a month of "working"... it's been confusing, really. A lot of new terms to understand, therefore a lot to read, and always thinking on where to apply it. Right now I'm finishing reading on Principal Components Analysis, PCA, and it's meaning (for more information, go here), Wikipedia rocks!). It's something that took more than two weeks of banging my head against the wall to understand, but it's worth it, because now I know that PCA can simplify my life in the near future.

Now that I'm full on this thing, as Dr. Lennox predicted, we have slowly changed course of where I'm going. It's ok though, I prepared myself for it and was expecting it: many other students that I've talked to have said that the original topic of research of a thesis is never the same as it is on when it's being written up. I'll probably wind up using this project as a training application of the terms that I'm reading upon and writing up something in System Modelling which is beginning to appeal to me. Coincidentally, that is what the bank is interested in.

I'm ok here... confused, but apparently is normal. I've never found comfort in confusion, there's no control there, but irony is now my way of life and it looks comforting =)

What if everything around you
isn't quite as it seems?
What if all the world you think you know
is an elaborate dream?
And if you look at your reflection,
is it all you wanted to be?
What if you could look right through the cracks?
Would you find yourself
find yourself afraid to see?

What if all the world's inside of your head
just creations of your own?
Your devils and your gods,
all the living and the dead,
and you really are alone.
You can live in this illusion.
You can choose to believe.
You keep looking but you can't find the woods,
while you're hiding in the trees.

Good song...