Honors Math

<p>We use the third edition of Spivak, so you will need to upgrade(the first edition was published in 1967, the second in 1980, and the third in 1994, and the preface notes significant changes among the editions).</p>

<p>Every problem set we have had has been directly from the book with one exception. On the current pset Barrett put two questions on the handout itself--they are not in the book.</p>

<p>I had forgotten one thing about psets. We are given two sets of problems each week: study problems and problems to turn in. We are given an ample number of study problems, and the problems we turn in often overlap the study problems; if one did every problem, I can see how he would spend twenty hours on homework every week.</p>

<p>Cool. How would you describe the level of the students in the class? Are there a lot of really talented kids, or is more made of people who just enjoy math?</p>

<p>I'd say the class has a mixture of both, but most fit into the latter.</p>

<p>It's hard to say who is talented and who is not, because the instructor drives the lecture, but there are obviously a few who are talented. Almost if not all of the students are Honors LSA, so the peer base is among UMich's best. I have heard a few muse that they were tricked into taking the class during orientation. Nevertheless, my impression is that some in the class liked the idea of taking UMich's hardest math, because they did well in math in high school, if you understand what I am saying.</p>

<p>For the first midterm, the median was a 70; it definitely followed a bell curve, and this was not because Barrett graded the tests to force a curve.</p>

<p>Honestly though, I think I'm not seeing many of the talented kids; i.e., they don't come to class! While a number of students did horribly on the first midterm, a good number scored very well.</p>

<p>Ironically, I said that psets were never more than five hours. This last one took me around seven.</p>

<p>But, yes, the level of student varies in the class. A student does not even need knowledge of calculus before entering the course, though I know there are those in class who are familiar with more advanced topics, e.g. ring theory for one.</p>

<p>edit: also, everything may be relative; supposedly no one got a grade lower than a B- last year. However, Barrett also said people did a lot better on the second midterm.</p>

<p>edit2: I feel as if I'm not giving the class enough credit. Around 1/4 - 1/3 of the class scored above 80 on the first midterm. Btw, the midterm was five questions in two hours, if that puts things into context (it was rough). In any case I wouldn't worry about the class being similar to high school math; math 295 has a much better base, both in terms of peers and actual material.</p>

<p>And the reason for the much longer pset was one of the two problems Barrett added this time; here's an example of one of the class's more challenging problems:</p>

<p> ext{Suppose that }x<em>1,\ldots,x</em>n ext{ are postive and that }m ext{ is an integer greater than 1. Show that }\sum^n<em>{i=1}x</em>i\le (n^{1-\frac{1}{m}}) imes\sqrt[m]{\sum^n<em>{i=1}x</em>i^m}</p>

<p>AoPS</a> TeXer</p>

<p>We are currently on the concavity chapter.</p>

<p>So I've worked on that problem for a bit.</p>

<p>It looks like a fairly direct application of Holder's Inequality. The case for m=1 is trivial, for m=2 it becomes Cauchy, which suggests Holder, or</p>

<p>\sum^n<em>{k=1} a</em>k b<em>k \leq (\sum^n</em>{k=1} a<em>k^p)^{1/p}(\sum^n</em>{k=1} b_k^q)^{1/q} </p>

<p>Putting a<em>k=1 and b</em>k=x_k for all k, k=1,2,...,n, and letting q=m and p=m/(m-1) we get</p>

<p>\sum^n<em>{k=1} x</em>k \leq (1+1+...+1)^{1/(m/(m-1))} (\sum^n_{k=1}x^m)^{1/m}</p>

<p>which is the desired inequality.</p>

<p>My question is why use the n^(1-1/m)? It seems very arbitrary.</p>

<p>EDIT - haha, no it doesn't. Way for me to forget Holder. 1/p+1/q must be 1, so it is necessary to put m/(m-1) for p if q will be m. Then 1/p+1/q = (m-1)/m+1/m = 1. Yay.</p>

<p>Jensen would be dissapointed.</p>

<p>Holder's Inequality is pretty high powered for this problem :)</p>

<p>But, I wasn't really playing fair either.</p>

<hr>

<p>Definitions of convexity:</p>

<p>Def 1: A function $f$ is convex on an interval, if for all $a$ and $b$ in the interval, the line segment joining $(a,f(a))$ and $(b,f(b))$ lies above the graph of $f$.</p>

<p>Def 2: A function $f$ is convex on an interval if for $a$, $x$, and $b$ in the interval with $a<x<b$ we have
$\frac{f(x)-f(a)}{x-a}<\frac{f(b)-f(a)}{b-a}$</p>

<p>One of the problems on the assignment:</p>

<ol>
<li>Show that $f$ is convex on an interval if and only if for all $x$ and $y$ in the interval we have
$$f(tx+(1-t)y) < t\cdot f(x)+(1-t)f(y), ext{ for }0<t<1.$$
(This is just a restatement of the definition, but a useful one.)</li>
</ol>

<p>One of the study problems:
9)
Let $p<em>1,\ldots,p</em>n$ be positive numbers with $\sum^n<em>{i=1}p</em>i=1$</p>

<p>(c): Prove Jensen's Inequality: If $f$ is convex, then</p>

<p>$f\left(\sum^n<em>{i=1} p</em>i x<em>i\right)\le \sum^n</em>{i=1} p<em>i f(x</em>i).$</p>

<p>Hint: Use Problem 4, noting that $p_n=1-t$</p>

<hr>

<p>If you manipulate the original expression a little, you'll see that you have a straightforward application of Jensen. The trick is proving Jensen, which takes a little work (there's a problem for you).</p>

<p>My</a> proof of Jensen for my own indulgence</p>

<p>Way to blow by all that though, haha.</p>

<p>I didn't check my TeX code this time, so there may be a problem or two, but I must take notes for my 295 midterm tomorrow!</p>

<p>Cool! I think I can do the problem with Jensen. </p>

<p>I did know Jensen previously, but this problem screamed Holder. I have the same induction proof (yay for inequalities class at mathcamp.)</p>

<p>Good luck on the midterm, and give make sure to indulge us with some problems =p.</p>

<p>Anyway, this thread has me convinced. I'm going to take the class next year.</p>

<p>just a quick note ....</p>

<p>The difficulty of the class is greatly overrated. Attend the lectures and make sure you have every def/thm/lemma as notes for the midterms/finals. There's probably a crazy curve, since a crazy number of people didn't show for the final (already failing? not sure).</p>

<p>< 5 hrs of homework a week => A for the semester.</p>

<p>If you like math, learn relatively quickly, and can take tests well (i.e., always seem to have fortuitous epiphanies), enroll 'cause Spivak's the bomb.</p>

<p>Any questions? Ask.</p>

<p>btw, it screamed Holder b/c you use Jensen to prove Holder :P</p>

<p>Wow sonar, you just revived a thread that's over a month old. Congrats And don't speak too soon.It appears we're classmates, and I'll see you in 296 (whoever you are), but to all viewers out there, 295 is VERY RARELY as sonar described. In most years, it's much more work, but the same very generous grades. This year was surprisingly easy, and that's mostly because Barrett taught it. If they let DeBacker or Conrad teach again...be prepared for hellish workloads.</p>

<p>if that is the case, my pardons.</p>

<p>Hehe...this thread is being revived again.</p>

<p>So...how's Spatzier? He's teaching 295 next year.</p>

<p>Also, a merit scholarship from the math department ensures that you're in the Honors Mathematics sequence, right? I believe so, but I'm just making double sure...</p>

<p>I know nothing of Spatzier as a teacher, but I do know he's the scholarship advisor for international students and as far as that goes, my friend thinks he's pretty good. Later on in the summer I'm helping with a class at MMSS, so I can ask around about him then.</p>

<p>For the Merit Scholarship, it's just required that you major in math in order to continue getting money, it's pretty much independent of the Honors math sequence (though most people getting money take it). At least the way it seemed when I went through the sequence was that anybody could sign up for the class if you asked for permission, a lot of people dropped the class early on, and at the end of the first semester we were basically told if you got a B+ that you weren't cut out for 296.</p>

<p>I have a feeling that next year's Honors Math Sequence is going to be like baby bear in that it's juuust right. Two years ago when I took it, Debacker made it a point to crush our souls in preparation for differential geometry/Conrad in 395-396. Last year, Barret and Burns seemed content to go straight out of the book and not really do much more. I think Spatzier is probably going to know that there's stuff that needs to be covered outside of the course text, but not go too overboard with it.</p>

<p>Yaaaaaaaaaaay! Michigan forum nerd playgorund returrns! We love you christine. Yeah, you don't have to be a scholarship student to take honors, you just have to ask permission, but be forewarned that about 50% make it out of 295 and about 50% of them make it out of 296. For dilksy's year i think it went from roughly 45 to 13 from 295 to 395, for my year it's going to be 50 to 15 I believe. But if you love math, you'll have no problem staying in. The people that drop are usually just the people that want to take a hard math class because it's hard and in high school some ignoramus told them they were good at math. Pretty much everyone that actually loves math stays. I got nothin to offer about spatzier, but i hope he's as dilksy predicts.</p>

<p>Yay! Much help! When I spoke with Debacker, he said that his classes usually start with about 40 people and end with 10..."Which is good." Oh yes, I love math...should be fun. Thanks!</p>

<p>Also Christine, assuming that's your name, you should probably seek out a certain sophomore math major next year that dilksy and I know who is known to bet a Stucci's ice cream (gourmet stuff...sounds good right now actually) that any given female won't make it to 395. He made the bet with one girl this year and won. If you're certain you've got what it takes to make it, I'd recommend cashing in on the free ice cream.</p>

<p>It's nice to know that college math boys are just as mature as high school math boys are just as mature as pre-pubescent math boys. Should I cash in on Stucci's ice cream? You know, I've never heard of Stucci's. Is it good?</p>

<p>I guess I should formulate 2 plans. Plan A: I make it. Plan B: I have to figure out some way to make this plan backfire on him.</p>

<p>Does this guy have a facebook or something?
I want to be his facebook friend. <a href="mailto:ckhwang@umich.com">ckhwang@umich.com</a></p>

<p>Goodness, have I been good at imitating a girl all this time...</p>

<p>In his defense, his bet was purely an exercise in expected value, not a result of some inherent feeling that girls suck at math. Last year 395/396 was comprised entirely of males and this coming year I believe it will be between 85% and 100% male.</p>

<p>Not only were we entirely male, but we were entirely single for most of the year (teacher included). There were two at the start with girlfriends, but one broke up with her and one dropped the class, then towards the end somebody finally broke the dry spell.</p>

<p>I'd say the ice cream is worth cashing in on. If only to watch this kid lose a bet. Well, I saw him lose the bet "I can hook up with any girl I want" rather frequently on the weekends, but only because every girl I mentioned was rendered ineligible.</p>

<p>I know. I understand that his bet was purely in exercise value...and I admit, I'm a tad bit of an anti-feminist myself. A little more than a tad bit, actually...</p>

<p>I'm not single, but I will be when I start UMichigan. Or tomorrow...depending on how Michigan versus Vanderbilt argument folds out. My boyfriend's going to Vanderbilt and I'm going to Michigan...each of us thinks our school is better. But Michigan's better. By a mile. So he can go to the southern frats and have his fun, get married when he's 20 like everyone else I know. I'm getting out of the South. And Michigan's better. (I'm just kidding about the fight determining the length of our relationship...I think.) Michigan is so much better.</p>

<p>So I sent a message to this guy. Waiting for him to reply... Stucci's isn't really expensive or anything, is it?</p>

<p>Oh good. Honors math majors still have time to do dumb things on weekends. That probably means that I'll be spending that extra time studying...I will stay on the track.</p>