Math-related Interview Questions - Page 3
Blogs > DTK-m2 |
raiame
United States421 Posts
| ||
Oracle
Canada411 Posts
On March 30 2011 01:19 ]343[ wrote: did you mean to say (4,5) is not a jam? or are numbers to the left the cars in front... or are your numbers inversely related to the speed. ok I'm confused lol. let a number represent speed and coordinates with higher index be in front (as if cars traveled right). so 5 4 3 2 1 is how many jams? 1? how about 2 3 1? does the 2 count as part of the same jam, since the 3-car is slowed to 1? Sorry i made an error in my post, i updated it. 5 4 3 2 1 is zero jams, since 5 travels faster than 4, faster than 3, etc. so none would be bottlenecked by another car 2 3 1 would be 1 jam, since 3 is bottlenecked by 2, but 1 cannot catch up to 2 to be bottlenecked. (i am reading the cars travelling left, not right, thats where u may be confused) 1 2 3 4 5 would be a single jam 1 3 2 would be a single jam + Show Spoiler [Hint] + my answer requires calculus/finite series | ||
Oracle
Canada411 Posts
A: "I have three daughters now. The product of their ages is 72. and the sum of their ages is equal to the amount of years we have known each other. B: "I still don't know how old they are" A: "My youngest daughter just started to play the piano" B: "Oh, my youngest is the same age!" How old are person A's daughters? + Show Spoiler [hint] + Ambiguity implies what? | ||
]343[
United States10328 Posts
On March 30 2011 02:15 Oracle wrote: 8) Two mathematicians who haven't seen each other in 15 years are talking. A: "I have three daughters now. The product of their ages is 72. and the sum of their ages is equal to the amount of years we have known each other. B: "I still don't know how old they are" A: "My youngest daughter just started to play the piano" B: "Oh, my youngest is the same age!" How old are person A's daughters? + Show Spoiler [hint] + Ambiguity implies what? 6*6*2 = 3*3*8 = 72, and both have sum 14; they're the only triples with product 72 that have the same sum (hence ambiguity). but since he has a distinct youngest daughter, she must be 2. ... except they can't have known each other for less than 15 years =.= but I'll assume that was extraneous information that happens to be false ;P | ||
YejinYejin
United States1053 Posts
On March 30 2011 01:26 raiame wrote: I wish all Jane Street cared about was math haha, got to final round and Tim told me I was too quiet and mumbled too much =[ You applied to Jane Street? How hard would you say the math was? | ||
Bibdy
United States3481 Posts
On March 30 2011 02:15 Oracle wrote: 8) Two mathematicians who haven't seen each other in 15 years are talking. A: "I have three daughters now. The product of their ages is 72. and the sum of their ages is equal to the amount of years we have known each other. B: "I still don't know how old they are" A: "My youngest daughter just started to play the piano" B: "Oh, my youngest is the same age!" How old are person A's daughters? + Show Spoiler [hint] + Ambiguity implies what? Did some Googling. Its an interesting problem. + Show Spoiler + The only knowledge you initially seem to have to go on is that the product of their ages is 72. The second clue seems meaningless. However, the final clue tells you in that one of the numbers is strictly smaller than the other two (which could just as easily be the same number as not). Since 72 isn't a perfect cube, then there's a limited number of answers it could be. a * b * c = 72 a + b + c = D Possible values are ============== a = 24, b = 3, c = 1, D = 28 a = 18, b = 4, c = 1, D = 23 a = 12, b = 6, c = 1, D = 19 a = 12, b = 3, c = 2, D = 17 a = 9, b = 8, c = 1, D = 18 a = 9, b = 4, c = 2, D = 15 a = 8, b = 3, c = 3, D = 14 a = 6, b = 6, c = 2, D = 14 a = 6, b = 4, c = 3, D = 13 Because of the first clue, a must be less than 15, and because of the second clue, D must be greater than 15, which leaves us with: a = 12, b = 6, c = 1, D = 19 a = 12, b = 3, c = 2, D = 17 a = 9, b = 8, c = 1, D = 18 So, the youngest daughter is either 1 or 2, but since a 1-year old is mostly still crawling around the floor, rather than playing piano, her age must be 2, which leaves the other daughter's ages as 3 and 12. Edit: + Show Spoiler + Although the poster above, ]343[, brings up a good point. Mathematician B would know the value of D, but the only reason to ask why they didn't know the answer after being given the second clue would be if there was ambiguity in the results, and that only occurs where D is equal to 14, so the only two outcomes are: a = 6, b = 6, c = 2, D = 14 a = 8, b = 3, c = 3, D = 14 Since there's a youngest, the answer has to be 6, 6, 2. It seems like these kinds of interview questions don't really have a purpose. If someone asked me about this problem in an interview now, I'd be able to solve it in 20 seconds. If they asked me 20 minutes ago, I'd just stumble around for 10 minutes and give up. All they reveal is how much brushing up someone did prior to the interview. | ||
Oracle
Canada411 Posts
On March 30 2011 02:45 ]343[ wrote: 6*6*2 = 3*3*8 = 72, and both have sum 14; they're the only triples with product 72 that have the same sum (hence ambiguity). but since he has a distinct youngest daughter, she must be 2. ... except they can't have known each other for less than 15 years =.= but I'll assume that was extraneous information that happens to be false ;P Sorry typo, my fault, i dont really proofread at work | ||
raiame
United States421 Posts
On March 30 2011 02:46 DTK-m2 wrote: You applied to Jane Street? How hard would you say the math was? Early rounds would be problems similar to those given here. Later rounds math wasn't easy enough that you could solve and you would just have to guess and make markets/confidence intervals and give thought processes/reasoning. Final round was mostly on decision making based on mathematical questions. | ||
bpgbcg
United States74 Posts
On March 29 2011 17:54 DTK-m2 wrote: I made AIME in high school and got an 11 ...what I thought you got a 6. Was this in a different year? Darn I can't stop thinking about the "jams" problem. Anyway, progress so far: + Show Spoiler + This is equivalent to the number of increasing subparts (i'm sure there's a more technical term) of a permutation of length at least 2. It's probably easier to include all increasing subparts and then subtract the expected number of increasing subparts of length 1. When do those occur? when a number has a higher number before it and a lower one after it (or just the relevant one if it's the first or last). Thus there are an expected 2/2+(n-2)/4=(n+2)/4 increasing subparts of length 1. Now to count the total number of increasing subparts. Since my combo is awful I will attempt to biject and probably fail, but interesting problem! BTW, does LaTEX work on TL? | ||
]343[
United States10328 Posts
On March 31 2011 10:49 bpgbcg wrote: ...what I thought you got a 6. Was this in a different year? Darn I can't stop thinking about the "jams" problem. Anyway, progress so far: + Show Spoiler + This is equivalent to the number of increasing subparts (i'm sure there's a more technical term) of a permutation of length at least 2. It's probably easier to include all increasing subparts and then subtract the expected number of increasing subparts of length 1. When do those occur? when a number has a higher number before it and a lower one after it (or just the relevant one if it's the first or last). Thus there are an expected 2/2+(n-2)/4=(n+2)/4 increasing subparts of length 1. Now to count the total number of increasing subparts. Since my combo is awful I will attempt to biject and probably fail, but interesting problem! BTW, does LaTEX work on TL? lolsup I've seen your username on AoPS. anyway LaTeX doesn't work, just BBCode. (why would LaTeX work on a non-math forum...?) though if you want, I'm sure you could use the TeXer and just include the images lol. also, I'm not sure what you're talking about in the "subparts" at all. please elaborate. | ||
YejinYejin
United States1053 Posts
On March 31 2011 10:49 bpgbcg wrote: ...what I thought you got a 6. Was this in a different year? Darn I can't stop thinking about the "jams" problem. Anyway, progress so far: + Show Spoiler + This is equivalent to the number of increasing subparts (i'm sure there's a more technical term) of a permutation of length at least 2. It's probably easier to include all increasing subparts and then subtract the expected number of increasing subparts of length 1. When do those occur? when a number has a higher number before it and a lower one after it (or just the relevant one if it's the first or last). Thus there are an expected 2/2+(n-2)/4=(n+2)/4 increasing subparts of length 1. Now to count the total number of increasing subparts. Since my combo is awful I will attempt to biject and probably fail, but interesting problem! BTW, does LaTEX work on TL? Ah fine, you caught me. Hi Ben. :D I got a six freshman year, and senior year I ALMOST got an 11. Like, I would have had it if I knew how to do addition. And simple multiplication. Or read English. My actual score was a 7? or an 8? One of those 2. | ||
]343[
United States10328 Posts
On March 31 2011 12:29 DTK-m2 wrote: Ah fine, you caught me. Hi Ben. :D I got a six freshman year, and senior year I ALMOST got an 11. Like, I would have had it if I knew how to do addition. And simple multiplication. Or read English. My actual score was a 7? or an 8? One of those 2. oh wait. he's... ben from... GDS?! oh that would explain the 14 loooool ok it all makes sense now! | ||
MilesTeg
France1271 Posts
On March 29 2011 21:40 Wonders wrote: Is it true that some of these firms test you on the raw speed of your mental arithmetic? It's a different kind of firm. Jane Street is more like a hedge fund as far as I know, while the speed math tests are typically in market making / arbitrage firms. In Europe/ Asia some examples of such firms are Optiver, All Options, Tibra... great career if you can pass the tests. In the US I don't know them as well, there are a lot of these firms in Chicago but for some reason they don't seem to have much presence internationally as far as I can tell. | ||
bpgbcg
United States74 Posts
So + Show Spoiler + If higher number=higher speed, each jam is started by a number for which the numbers on either side are both higher than it. (If the one on the left was lower it would not be the beginning of a jam, if the one on the right was lower it would not cause a jam at all.) So what's the probability for this to happen for any given number? For the n-2 middle numbers, the probability that out of any 3 adjacent ones the middle is the smallest is clearly 1/3. However, for the one on the right side it cannot start a jam, whereas for the one on the left the probability is 1/2. Thus we have (n-2)/3+1/2=(2n-1)/6. Also ]343[ do I know you? | ||
Oracle
Canada411 Posts
The arithmetic solution which I got was + Show Spoiler + 1 + 1/2 + ... + 1/(n-1) + 1/n I didn't give the explanation but sometimes it helps to know the answer and work backwards. | ||
bpgbcg
United States74 Posts
+ Show Spoiler + If the slowest car is the kth car from the back it creates a jam behind it. In addition if f(n) is the number of expected jams, there are f(n-k) expected more jams in front of the slowest car. Since each of these is equally probable, we have f(n)=(f(n-1)+f(n-2)+...+f(1))/n+1. Thus nf(n)=f(n-1)+...+f(1)+n. Thus nf(n)-(n-1)f(n-1)=f(n-1)+1, so f(n)=f(n-1)+1/n. Induction trivially yields f(n)=H_n. | ||
womiebra
1 Post
With that simplification I can confirm that answer. Here is how i got it: + Show Spoiler + X_i = 1 if i-th slowest car is blocked by a jam, 0 otherwise Y - number of jams (including 'single-car jams') The following holds (P denotes probability): P(X_i=1) = 1 - 1/i explanation: look at subset {1,2,...,i-1} of i slowest cars. The chance that i-th car has no slower car anywhere in front equals to the chance that i is at the first position of i-permutation, which is 1/i. Observe that (E denotes expected value): Y = N - (sum of EX_i) Then the answer to the simplified problem is: EY = N - (sum of EX_i) = N - (N - 1 - 1/2 - ... - 1/N) = 1+1/2+...+1/N = H_N Would be interesting to know the answer for the original problem though | ||
MoonBear
Straight outta Johto18973 Posts
As for the standard of math. If you want to get into finance, having done further mathematics at secondary school is a must. Having done mathematics or quantitative subjects at university is recommended. While you don't necessarily have to have done finance at university, a lack of mathematical background will set you back. This is especially important for the more quantitative-finance orientated positions, such as say Financial and Audit Consulting at one of the Big Four. Different companies will also ask different styles of questions. You'd probably want to do some research or talk to previous applicants about what the company is looking for. Different departments in different countries will also have different standards on what they want because the recruiters all have their own preferences so check beforehand if you can. Ultimately, trying to read up on every brainteaser in existence probably won't be a great strategy as there's too many to learn. You'll probably have to do psychometric testing as well and you don't want to perform poorly on them. Those matter a lot because if you don't make the cut, they just reject you then and there. I'd also recommend knowing how to do mental math fast. For example, knowing the tricks to multiplying any two digit number by 11, or converting decimals into fractions. As for the questions themselves: Question 1: + Show Spoiler + The answer is 7. Divide your horses into 5 groups of 5. Race each group of 5. Let us call each group A, B, C, D, E with and the rank the horses 1, 2, 3, 4, 5. We then have: A1, A2, A3, A4, A5 B1, B2, B3, B4, B5 C1, C2, C3, C4, C5 D1, D2, D3, D4, D5 E1, E2, E3, E4, E5 Now we race A1, B1, C1, D1, E1. Let's just say that the final result has A being the fastest, B coming second, C third, etc. Since these are just labels, you can relabel them with something else if you want. We can eliminate everything in group D and E for obvious reasons. We also eliminate C2 to C5 as we can always form a group of 3 faster horses (e.g. A1, B1, C1). The same logic allows us to eliminate B3 to B5 and A4 and A5. We have A1, A2, A3, B1, B2, C1 left. We know A1 is the fastest horse of them all. So race the remaining 5 and choose the two fastest from the rest. Question 2: + Show Spoiler + I don't actually understand this question. Is each string only allowed through 1 hole? Question 3: + Show Spoiler + The fact they don't burn uniformly doesn't matter. Light string 1 at both ends at the same time and light string 2 at one end. String 1 will fully burn in 30 minutes. String 2 will be half burnt. When String 1 fully burns, light the non-lit end of string 2. String 2 will now burn for 15 more minutes on top of the 30 min it has already burnt. Question 4: + Show Spoiler + The answer on Page 1 was correct. Question 5: + Show Spoiler + The solution on Page 2 works, but you probably won't get out of jail until the universe ends because the amount of time it would take for everyone to be randomly selected and have visited the room and then have one person go back would take far too long. I believe the most elegant solution would be this one which also covers every possible variation of the problem. Question 6: + Show Spoiler + This a pretty well known question with many explanatory solutions so I won't type one up here. | ||
jvanbev1
United States1 Post
| ||
| ||