The midterm results are in, and they're not pretty. With a mean of 33 and a standard deviation of 18, we can safely say that this is not a normal distribution (assuming non-negative scores). It's been a while since I've taken an exam with such a low mean; it almost brings back nostalgia for the Engineering Math sequence. As I waited for the TA to find my test, a slew of 7s, 15s and 18s flashed by, and I began to silently hyperventilate. You know things are not good when your score < your age. Moreover, the stakes are higher now than in undergrad. I not only have my transcript, but $2,500 on the line, since the Fed won't reimburse me for tuition unless I get a C or above.
I knew this wasn't going to be an easy test, but I severely underestimated Prof Mykland's exam-writing abilities. Perhaps I should have known better when he related this anecdote on his PhD class: "I offered my class of ten students automatic As if they showed up for the exam...and 2-3 of them decided they'd rather skip it." When he announced that he wasn't allowing calculators, I thought that simply meant computations would be more tedious. In reality, that meant the exam would be completely theoretical.
In retrospect, perhaps I should have paced myself a bit better. We had 3 hours to complete 3 questions...and it wasn't nearly enough time for me. I ended up getting full credit for the first question, but by the time the 3rd question rolled around, I was so frustrated that I didn't even bother attempting to get partial credit, I just threw out the last page.
Anyways, I managed to beat the mean, though not by a healthy amount. Mykland reassured us that the vast majority of students receive As and Bs overall, and that if we did better on the final, those results would count in full. Never mind that the material covered on the final will be much harder. Mykland did say that he'd try to tone it down a bit for the final, since usually he gets means of "at least 40-50."
Student: "I understand that you don't want to write an exam that's too easy, but how am I supposed to feel about getting a 33 when I'm interviewing for jobs? If someone were to ask me about martingales, I would not say with confidence that I understand them very well."
Mykland: "Well, as evidenced by what's happening on Wall Street, I'd say a great many people don't understand them very well...if anything, I don't want to give anyone the illusion of understanding. The more you know, the more you realize you don't know."
Sigh. Okay, I can figure out the Itô integral of a deterministic integrand in the next three weeks...
Showing posts with label mathematics. Show all posts
Showing posts with label mathematics. Show all posts
Monday, November 17, 2008
Thursday, January 17, 2008
All-Nighter at Club Fed
Since I like to be masochistic, I decided to take a class this quarter on top of working full-time. Well, that wasn't exactly the reason, but I figured I should dip my toes back into coursework to strengthen my transcript for grad school apps. More importantly, after being out of the classroom for 6 months, I need to reassess whether I want to go back for a few more years of fretting about exams, or just stay in the "real" world. Luckily, the Fed offers tuition reimbursement for "related" classes that improve human capital, which loosely covers all kinds of business, economics and math classes. Have I mentioned lately how much my job rocks?
The first class I'd wanted to take was Intro to Stochastic Processes, but Prof Lalley pretty much told me it was geared toward PhD students and that I should take something else. So, I settled on an "Introduction to Probability Models" class at the U of C (University of Chicago, in local parlance), which looked like it'd be manageable. It also covers stochastic processes, but with a less theoretical treatment and more emphasis on applications. And as we all know, applications = easy (relatively). Finally, in my smartest move yet, I was able to convince Katherine to take the class with me, so that I wouldn't be studying and doing problem sets entirely on my own.
The class is cross-listed as a graduate course, so it's filled with masters statistics students and senior math majors. To be honest, I'm more afraid of the senior math majors, many of whom have serious bowl cuts going on. When I first walked into the class, the room was filled with Asians. Oh wait, there were two white people, and they looked Russian. By the end of week one, we were 3 chapters and 200 pages into the book. The problem sets are due on Thursdays, and I spent the weekend reading the book, and did a first pass of the problem set on Mon night. At this point, I started to worry, and told Katherine that we had to work on this problem set Tues night. We stayed at the Fed until 11 pm, and at that point we'd only covered 3 out of the 7 problems...and the questions only get harder as you go on.
Wed night was even better--we spent the entire night at the Fed. That's right, we went to work at 9 in the morning, stayed up all night doing the first problem set of the quarter, went to class the next morning, and then came back for another 8 hours of "real" work. And because it was the week before an FOMC meeting, we were swamped with work. It wasn't pretty.
On the other hand, I've never been more comfortable with probability in my life. We were in an office with a whiteboard wall, which was covered with crazy scribblings and mathematical notation by the end. I felt really happy with our progress. Between the two of us, we were coming up with all sorts of re-indexing methods, recognizing obscure probability distributions, and transforming ugly algebraic messes into compact sums like no other. There's no way I would've been able to come up with all that on my own. And it was really satisfying to know that I could still think and analyze on a higher-order level.
As much fun as that was, I'm hoping that subsequent problem sets lighten up a bit, because I can't afford to pull an all-nighter every week. Back in my undergrad days, I could easily turn in projects and then sleep for the rest of the day, but that's not an option any more now that I have a job. Erg.
The first class I'd wanted to take was Intro to Stochastic Processes, but Prof Lalley pretty much told me it was geared toward PhD students and that I should take something else. So, I settled on an "Introduction to Probability Models" class at the U of C (University of Chicago, in local parlance), which looked like it'd be manageable. It also covers stochastic processes, but with a less theoretical treatment and more emphasis on applications. And as we all know, applications = easy (relatively). Finally, in my smartest move yet, I was able to convince Katherine to take the class with me, so that I wouldn't be studying and doing problem sets entirely on my own.
The class is cross-listed as a graduate course, so it's filled with masters statistics students and senior math majors. To be honest, I'm more afraid of the senior math majors, many of whom have serious bowl cuts going on. When I first walked into the class, the room was filled with Asians. Oh wait, there were two white people, and they looked Russian. By the end of week one, we were 3 chapters and 200 pages into the book. The problem sets are due on Thursdays, and I spent the weekend reading the book, and did a first pass of the problem set on Mon night. At this point, I started to worry, and told Katherine that we had to work on this problem set Tues night. We stayed at the Fed until 11 pm, and at that point we'd only covered 3 out of the 7 problems...and the questions only get harder as you go on.
Wed night was even better--we spent the entire night at the Fed. That's right, we went to work at 9 in the morning, stayed up all night doing the first problem set of the quarter, went to class the next morning, and then came back for another 8 hours of "real" work. And because it was the week before an FOMC meeting, we were swamped with work. It wasn't pretty.
On the other hand, I've never been more comfortable with probability in my life. We were in an office with a whiteboard wall, which was covered with crazy scribblings and mathematical notation by the end. I felt really happy with our progress. Between the two of us, we were coming up with all sorts of re-indexing methods, recognizing obscure probability distributions, and transforming ugly algebraic messes into compact sums like no other. There's no way I would've been able to come up with all that on my own. And it was really satisfying to know that I could still think and analyze on a higher-order level.
As much fun as that was, I'm hoping that subsequent problem sets lighten up a bit, because I can't afford to pull an all-nighter every week. Back in my undergrad days, I could easily turn in projects and then sleep for the rest of the day, but that's not an option any more now that I have a job. Erg.
Subscribe to:
Posts (Atom)
Posts
