Saket Vora

So ’settling’ into the semester means that much of my time is spent working on problem sets for EE 214 (analog circuits) and EE 278 (statistical signal processing). Needless to say, I enjoy the 214 homework better. If engineering classes lie on a continuum of practical-oriented to theoretical-oriented, these two classes are on opposite ends. Our analog circuits class is deep into the practical — we are introduced to the standard long-channel equations describing MOSFETs, then spend an entire lecture examining why they are essentially impractical for designing real circuits. The approach we are learning — and use in the homeworks — is a combination of equation and simulation results that come directly from the process we’ll be using for the project. Statistical signal processing, however, carries itself as a math course. Of course, it is indeed math based, but when the student is presented with a theorem or a rule, the supplied course reader (no textbook is assigned) goes on to present proofs of why said theorem is true. The few and far between examples are given only as ‘exaggerated examples, not practical ones’ (this was heard in class, no less). The homeworks are where we predictably learn the most, because they are nearly all practical problems. As you can see, there is a big discrepancy there.

I’ve also acknowledged that for the first time, I need to visit office hours of the TA or professor in order to efficiently complete the homework assignments. In a typical week, I will spend an average of 4 hours each day Mon-Wed. working on 278 homework (due Thursday), then about the same amount of time Thursday to Sunday working on 214 homework (due Monday). I have found a study group for each class, and it has helped out tremendously.

Amusing aside from today’s 278 study session: we are doing estimation with multiple random variables that have probability density functions. These can be expressed as graphs. I used a graphical approach to one of the problems, because the mathematical notation required to precisely define the graphs were cumbersome to me. I was showing this approach to my study partners, showing how I set up an equation using pictures instead of variables. At one point, I had one ‘graph’ being divided by another ‘graph’. We joked about how ‘hand-wavy’ this was. Later, in the office hour session, the TA was walking through the same problem. She drew graphs on the board and remarked “so now we divide this graph here by that graph”. I smiled — and my partners were just shaking their heads.