Advice on how to get faster?

[Devin Jeanpierre]

Evening section/Lash-Miller person here. I did not come close to finishing the term test. I knew how to solve every problem, given enough time, but that's not really good enough.

It's not a very specific question, but does anyone have any advice for how to get faster at this for the midterm? Like, uh, maybe it would help if you said how much time you spend per week doing questions?

I'll find out if there were any technique differences when others post their answers to the problems. Like, I know I worked faster by using the in-class version of undetermined coefficients, but maybe I missed out on something when I missed the lecture that covered the wronskian... I don't know.

[Brian Bi]

I talked to a few people at random. Based on my sample, I believe that very few people were able to finish.

[Jason Hamilton]

I couldn't finish everything either, I had to leave a few integrals unsolved to finish the questions in time. 45min for this test was simply not enough time.

[Devin Jeanpierre]

There's not finishing and then there's not finishing.

I completed #1, #3, and #2a, but #1 had a mistake I couldn't find. I did not get anywhere (other than to state the roots) with #4, and I was in the middle of a long (and probably erroneous) computation for #2b.

Overall, not feeling so good about my grade. The way I figure it, in the best case I get maybe a 60%, and that's if people are nice marking -- but it's more likely I'll get a 40%, considering how harshly the quizzes were graded with regards to computation errors.

[Alexander Jankowski]

I am in the same position. I did no homework and relied purely on theory (as opposed to experience) in my solutions. Any of those questions are easily solvable in an adequate amount of time, but when time is as constrained as it was on the test, knowledge of shortcuts is the only way to score a flawless victory.

[Devin Jeanpierre]

Haha, are you another theory dude?

This is my first "computation" course. It's taking getting used to!

[Alexander Jankowski]

Yes, I prefer theory to computation. Unfortunately, I've not yet had the pleasure of taking a purely theoretical course. (I was in engineering.)

[Victor Ivrii]

First, we are dealing here with formal calculations (i.e. formulae) rather than numerical computations.

Second, good theory can speed up any calculations: in many problems you can observe that due to symmetry, parity etc some coefficients must be $0$. Or that some coordinate systems are preferable.

BTW, if you look at Math Journals you observe that many theoretical math papers look rather different: some of them have short and fewer formulae and other have more numerous and much longer. It depends on the field of math. Analysis (real or complex) and especially Differential equations (ordinary or partial) are formula-intensive.

[Rudolf-Harri Oberg]

I solved nearly every homework problem, I knew the theory well. Yet, I ran into computational difficulties and were not able to fully finish questions (i.e differentiate enough times to solve for constants, evaluate integrals etc)

From the given feedback so far and according to my personal experience, the message is clear: the questions selected for the test were not student-friendly or even unsuitable for a test of such a format (limited space, tight time constraints).

I feel somewhat deceived by the staff, instructors. Someone had not prepared for that Wednesday night and in my view it was not the students.

I hope the next test goes better.

[Victor Ivrii]

1) As discussion is not mathematical I moved it to another board.

2) First, this test was an almost exact copy of the test of 1 y.a. which is available (and not very different from the previous years).

Second, this test was printed on 4 double sided pages so for each problem you had almost complete double-sided page. And everyone could ask for an extra paper (we brought it, albeit we could forget to announce it--I was too busy with the list to notice anything).

3) Important: TT1 will be graded by Craig Sinnamon, and he will have an extra office hours for you to get your test and discuss with him. Since MidTerm which costs twice as much as TT1 will be based on the same material + Chapter 7 you definitely should use this opportunity. Recall:

February 25, Mon, 18:30—20:30 Craig Sinnamon in BA6283
February 26, Tue, 15:00—16:0 Craig Sinnamon in BA6283 and 17:00—20:00 in BA6283
February 28, Thu, 18:30—20:30 Craig Sinnamon in BA6283
March 01,       Fri, 11:00—12:00 Craig Sinnamon in BA6283 and 16:00—20:00 in BA6283

4) Those who missed Test please bring Doctor Notes etc to Prof Milman during his office hours.

5) On weekend I will place several more theoretical bonus problems (related to one another)

6) If you have some concerns please try first talk to instructors.

[Jeong Yeon Yook]

1. those people clearly did not mean numerical computations. We can tell by the context even if the use of the term was not the most correct. If it were numerical computations we would be writing algorithms instead of doing them by hand.

2.we did not discuss much about coordinate change or symmetry so far in this course other than straight calculations. so i wouldn't think that we are responsible for what the course was not intended for or what was never mentioned.

3. for math journals, although things are complicated they obviously did't have 45 minutes to write the journal in the exam center with other people staring down at them and counting down time under pressure.

and it would be reasonable to give students more time than what the professor needs to solve the questions since the professor is more experienced and better in the material while students are new to this and learning it for the first time.

[Victor Ivrii]

2.we did not discuss much about coordinate change or symmetry so far in this course other than straight calculations. so i wouldn't think that we are responsible for what the course was not intended for or what was never mentioned.

I just brought different examples, partially from Calculus, from Chapter 7 of our class, from APM 346 some of you take. Just to illustrate that calculations by hand (or computations by computer) are intertwined with theory.

3. for math journals, although things are complicated they obviously did't have 45 minutes to write the journal in the exam center with other people staring down at them and counting down time under pressure.

Again: to show that formulas and calculations are not opposed to the theory.

and it would be reasonable to give students more time than what the professor needs to solve the questions since the professor is more experienced and better in the material while students are new to this and learning it for the first time.

How do you know how much time we spend to solve these problems? Actually it takes much more time to invent the problem than to solve it.