/sci/ - Science & Math » Thread #14425700

129KiB, 638x777, Screenshot_1.jpg

View Same Google iqdb SauceNAO

A probabilty problem, you can do this /sci/

Anonymous Mon 25 Apr 00:46:26 2022 No.14425700 View Reply Original Report

Quoted By: >>14425847 >>14426030 >>14426588 >>14427289

Who is correct? (Not a homework, just a message board discussing math textbooks prohibited in Florida).

Anonymous

Anonymous Mon 25 Apr 2022 01:55:06 No.14425847 Report

Quoted By:

>>14425700
So this is the famous CRT math I've been hearing about.

Anonymous

Anonymous Mon 25 Apr 2022 02:14:37 No.14425875 Report

Quoted By: >>14425885 >>14425895 >>14425904 >>14431862

>>14425779
On the assumption that the variables are independent, you would observe 10% of the 500 that are armed to also be high on drugs, in the same fashion those 50 would also be 5% of the 1000 that are high on drug that are also armed.
>>14425830
7.5% for 2 successes of 100 trials with p=0,005, yes
That's not what is being asked here tho.

[[0.08340284 0.01417848]
[0.01417848 0.04378649]] is the covariance matrix of a 100 sample observation expressed as a binary Yes/No that conforms to the problem (same population shares, but 2% of the samples have both variables set to 1)

Anonymous

Anonymous Mon 25 Apr 2022 02:20:18 No.14425885 Report

Quoted By:

>>14425875
wow, I am retarded afterall

[[0.09090909 0.01515152]
[0.01515152 0.0479798 ]]

Anonymous

Anonymous Mon 25 Apr 2022 02:24:05 No.14425895 Report

Quoted By:

>>14425875
I have no idea what you're on about but my 8% figure should be close to the true value.

Anonymous

Anonymous Mon 25 Apr 2022 02:29:02 No.14425904 Report

Quoted By: >>14425943

>>14425875
>7.5% for 2 successes of 100 trials with p=0,005, yes
What are you talking about? 2 or more successes out of 5 trials with p=0.1, except that allows for the possibility of picking the same guy twice, so it overestimates just a little bit. It's ~8% in any case.

Anonymous

Anonymous Mon 25 Apr 2022 02:47:28 No.14425943 Report

Quoted By: >>14425997

>>14425904
I guess I'm retard as well. That doesn't scale linearly at all. With 10,000 people, that's 200 out of 500 trials which really is basically zero.

Anonymous

Anonymous Mon 25 Apr 2022 03:09:07 No.14425997 Report

Quoted By: >>14426025

>>14425943
that's ok, i've been trying to solve this 3 times, and got different answers because it is a trick problem or i am retarded. i guess the missing bit is if the variables are independent or not. i am pretty sure it is a tongue in cheek question otherwise it wouldn't be rejected by the florida board of education. perhaps the hint is if you are armed you are mostly likely a drug addict. of course alluding to a certain demographics. i don't care about the social aspect but the problem is interesting.

Anonymous

Anonymous Mon 25 Apr 2022 03:26:40 No.14426025 Report

Quoted By:

>>14425997
The only answer that makes sense is 0%, because no sample size is given and it's asking for exactly 2%, not at least 2% or some range around it, and there is no mention of the sample size or independece of the events.

Anonymous

Anonymous Mon 25 Apr 2022 03:29:54 No.14426030 Report

Quoted By:

>>14425700
50/50. either 2% of them are both on drugs and armed, or not.

Anonymous

Anonymous Mon 25 Apr 2022 03:30:41 No.14426032 Report

Quoted By: >>14426058 >>14431862

>>14425779
>But they are probably not, there is probably a significant correlation between being high and armed. maybe that's the whole point?
If they're dependent then this question is pointless. Is this a math problem or not bro?
>Both the first reply takes a sample of 10000
Taking a huge sample size applies the CLT which I referenced in my first post.
>But it is 5% of the entire sample, not 5% of those who are high on drugs. 5% of 10000 would be 500 not 50.
You need to relearn probability basics.

Anonymous

Anonymous Mon 25 Apr 2022 03:44:26 No.14426058 Report

Quoted By: >>14426074

>>14426032
>You need to relearn probability basics.
I don't understand. If you have 10k people. And you take 10% of 10k, that's 1000 who are on drugs. Then you know that 5% are armed. NOT those who are on drugs, but 5% of the sample. Perhaps totally different people. Not related to the 10% on drugs. 5% of 10k is 500. What is wrong with this reasoning?

Anonymous

Anonymous Mon 25 Apr 2022 03:52:15 No.14426074 Report

Quoted By:

>>14426058
Using the sample is skewing your thoughts to the discrete instead of continuous which it is in reality.
Because we're assuming that the two qualities are independent in terms of whether or not someone has drugs, there's no difference between 100 armed and 100 unarmed, it's still expected to be 5% of them.

Anonymous

Anonymous Mon 25 Apr 2022 04:12:40 No.14426104 Report

Quoted By: >>14426379

Plz faggots try harder. Its real problem but banned only because of racial sensitivity shit. Start to think harder so i dont have to

Anonymous

Anonymous Mon 25 Apr 2022 04:29:19 No.14426139 Report

Quoted By:

You should do gun crimes and drugs, jobless scum you will be. Simple question no answer

Anonymous

Anonymous Mon 25 Apr 2022 06:54:19 No.14426379 Report

Quoted By: >>14427260 >>14431862

>>14426104
It's zero because the problem asks the chance it's exactly 2%...not at least 2% or some range....exactly 2%. Since no sample sizes are given you can only assume continuous distributions.

Anonymous

Anonymous Mon 25 Apr 2022 07:19:00 No.14426414 Report

Quoted By: >>14426475 >>14431862

>>14425735
1st replier arrived at the corrct answer with bad logic...the 2nd one...not sure where he gets less than 1 in a trillion but that's wrong because it is exactly zero. In fact you dont even need the 5%/10% numbers to know its zero.

Anonymous

Anonymous Mon 25 Apr 2022 08:12:55 No.14426468 Report

Quoted By: >>14427701 >>14431862

>>14425735
that's not what the question is really asking. What is the probability of at least 2% being high and armed?

Anonymous

Anonymous Mon 25 Apr 2022 08:15:01 No.14426475 Report

Quoted By: >>14427249

>>14426414
>it is exactly zero.
No, it isn't.

Anonymous

Anonymous Mon 25 Apr 2022 09:07:13 No.14426588 Report

Quoted By: >>14426591 >>14428502 >>14431862

>>14425700
5% of S are A
10% of S are D

What is the probability 2% of S are both A and D?

Anonymous

Anonymous Mon 25 Apr 2022 09:08:35 No.14426591 Report

Quoted By: >>14426638

>>14426588
Well? What is it?

Anonymous

Anonymous Mon 25 Apr 2022 09:30:48 No.14426638 Report

Quoted By: >>14426643

>>14426591
there isn't enough information to determine it

Anonymous

Anonymous Mon 25 Apr 2022 09:32:10 No.14426643 Report

Quoted By: >>14427063 >>14428396 >>14428502

>>14426638
What information are you lacking?

Anonymous

Anonymous Mon 25 Apr 2022 13:59:30 No.14427063 Report

Quoted By: >>14427267 >>14428406 >>14431862

>>14426643
Not him, I think there should be a correlation between A and D to make the problem more interesting. Such as it is known that if you pick A there is a 70% chance you also pick D.
For example: it is generally known that 70% of diners order tater tots as a side if they order a hamburger. Now we have a sample of diners, say 10000. 5% ordered tater tots. 10% ordered a hamburger. What's the probability that 2% ordered both? (They could have tater tots with any other order).

Anonymous

Anonymous Mon 25 Apr 2022 15:26:29 No.14427249 Report

Quoted By: >>14427252

>>14426475
Come back when you're out of high school

Anonymous

Anonymous Mon 25 Apr 2022 15:27:44 No.14427252 Report

Quoted By: >>14427263

>>14427249
It's not zero for any finite number of people. Dilate.

Anonymous

Anonymous Mon 25 Apr 2022 15:29:25 No.14427260 Report

Quoted By: >>14427701

>>14426379
Of course it means OVER 2% NOW FUCKING CALC

Anonymous

Anonymous Mon 25 Apr 2022 15:31:04 No.14427263 Report

Quoted By: >>14427268

>>14427252
It's zero for any odd number of people..and again no sample size given.

Anonymous

Anonymous Mon 25 Apr 2022 15:31:47 No.14427267 Report

Quoted By: >>14427275

>>14427063
Calc you stupid moron dont cry. Lets assume they are independed things. Let say you do lottery 1:10000 and keno 1:100 and bet some horses 1:50 at toto. What are the odds that those who won in one game wins another

Anonymous

Anonymous Mon 25 Apr 2022 15:31:56 No.14427268 Report

Quoted By: >>14427290 >>14427701

>>14427263
> no sample size given.
Sample size = N. Protip: you will deflect because you're too dumb to solve it. :^)

Anonymous

Anonymous Mon 25 Apr 2022 15:34:52 No.14427275 Report

Quoted By:

>>14427267
...cont and what are the odds thet given 100 people 2 of them wins all the games

Anonymous

Anonymous Mon 25 Apr 2022 15:35:22 No.14427276 Report

Quoted By: >>14427283

I got 3.33% chance that '2% or more' were both. With the assumption they're independent etc etc.

Anonymous

Anonymous Mon 25 Apr 2022 15:41:49 No.14427283 Report

Quoted By: >>14427301

>>14427276
Show work thank you

Anonymous

Anonymous Mon 25 Apr 2022 15:43:07 No.14427289 Report

Quoted By:

>>14425700
>Assuming both variables (guns, drugs) are normally distributed in the population
"Hello officer how can I help you"
"I've pulled you over for speeding, do you have any weapons in the car?"
"Just the average amount"
"Excuse me, sir?"
"I'm only averagely armed, in fact I'm probably a standard deviation less armed than the median (drugs guns)"

Anonymous

Anonymous Mon 25 Apr 2022 15:43:32 No.14427290 Report

Quoted By: >>14427384

>>14427268
You were fatally incorrect in your previous statement so you are disregarded.

Anonymous

Anonymous Mon 25 Apr 2022 15:45:53 No.14427295 Report

Quoted By: >>14427299

Lottery ticket wins independently 5/1000. 100 000 buys one ticket. What is odds of that 2000 of them won?

Anonymous

Anonymous Mon 25 Apr 2022 15:46:56 No.14427299 Report

Quoted By:

>>14427295
I mean over 2000 sht shit shit

Anonymous

Anonymous Mon 25 Apr 2022 15:47:37 No.14427301 Report

Quoted By: >>14430422 >>14431862

>>14427283
I did it geometrically. Take a line segment of length 100 and mark a point 5 from the left end. The interval left of that point represents Drugs. Now you're going to place another interval of length 10 randomly on the line. If the overlap is 2 or longer, that's a success.

For the overlap to be 2 or more, the left edge of the new interval must be at the point 3 from the end, or further left. What's the probability of that? Well the space where the left edge of the interval can go has length 90, and the distribution is uniform, so the probability is 3/90

Anonymous

Anonymous Mon 25 Apr 2022 16:24:37 No.14427384 Report

Quoted By:

>>14427290
>i have a low iq

Anonymous

Anonymous Mon 25 Apr 2022 17:55:58 No.14427701 Report

Quoted By: >>14428353 >>14430429 >>14431862

>>14427260
>>14427268
>>14426468
It's still 0 bro. As n goes to infinity the cdf of the binomial converges to 1 when x > p ~ 0.02 > 0.005, which means the probability of it being greater is 1 - 1 = 0.

https://math.stackexchange.com/questions/2364979/finding-the-limit-as-k-tends-to-infinity-of-this-sum

Anonymous

Anonymous Mon 25 Apr 2022 21:37:10 No.14428353 Report

Quoted By: >>14428880

>>14427701
>>14427701
how the hell is it zero from a real life perspective? you have quite a bit of armed people and some high people. why is there zero probability that at least a certain number of them is both high and armed? it is obviously the case irl.

Anonymous

Anonymous Mon 25 Apr 2022 21:49:41 No.14428396 Report

Quoted By:

>>14426643
the lacking information is how much of S are A and D simultaneously

Anonymous

Anonymous Mon 25 Apr 2022 21:52:04 No.14428406 Report

Quoted By: >>14431862

>>14427063
>I think there should be a correlation between A and D to make the problem more interesting
the issue is that the question assumes that correlation without having any evidence for it

just make a table and you can see you can have 3 different cases of S, A and D (A and D don't intersect, A and D intersect and infinitely many points between 0% and 15%, A is inside D)

Anonymous

Anonymous Mon 25 Apr 2022 21:59:21 No.14428502 Report

Quoted By:

>>14426643
>>14426588
ok so the probability has to be something lower than 1/42 and higher than 0, that's your answer (since I don't know if 12.123% is considered a possibility too)

Anonymous

Anonymous Tue 26 Apr 2022 00:15:09 No.14428880 Report

Quoted By: >>14429011 >>14431862

>>14428353
To be clear it's not the proportion of people in the population that are high and armed. It is the probability that 2% or greater of the population is high and armed. When only 0.5% of the population are expected to be high and armed the probability of >2% occurring is tiny, and goes to 0 when n goes to infinity.

Like I said in a previous reply, if you knew the exact amount of spring breakers then a nonzero probability exists for when exactly 2% occurs.

Anonymous

Anonymous Tue 26 Apr 2022 00:58:13 No.14429011 Report

Quoted By:

>>14428880
OK, how would you re-word the problem so the answer would be a large positive number, say 70% probability that at least 2% of the spring breakers (a sample of 10k) are high and armed?

Anonymous

Anonymous Tue 26 Apr 2022 14:02:08 No.14430422 Report

Quoted By: >>14430436

>>14427301
So, is this the correct answer?

Anonymous

Anonymous Tue 26 Apr 2022 14:06:25 No.14430429 Report

Quoted By:

>>14427701
>As n goes to infinity
I didn't say N goes to infinity. I asked for the exact value given N. Just admit you're too stupid to find it.

Anonymous

Anonymous Tue 26 Apr 2022 14:08:38 No.14430436 Report

Quoted By: >>14431862

>>14430422
Firstly: no, because, it's wrong even for the values chosen, secondly: no, because the answer changes drastically as the number of people increases.

Anonymous

Anonymous Tue 26 Apr 2022 22:40:13 No.14431849 Report

Quoted By: >>14431862

OK, how about this correction:
10k people arrived in Miami. 5% were armed. 10% were high. Historically there is a 70% correlation between being armed and also being on drugs for this demographic. What is the probability that at least 2% of the group are both high and armed?

Anonymous

Anonymous Tue 26 Apr 2022 22:45:50 No.14431862 Report

Quoted By:

>>14425735
>>14425779
>>14425830
>>14425875
>>14426032
>>14426379
>>14426414
>>14426468
>>14426588
>>14427063
>>14427301
>>14427701
>>14428406
>>14428880
>>14430436
>>14431849
OK how about this correction:
10k people arrived in Miami. 5% were high and 10% were armed. It is known that historically there is 70% correlation between being armed and being on drugs for this demographic. What is the probability that at least 2% of the group are both high and armed?

Capcode	All Only User Posts Only Moderator Posts Only Admin Posts Only Developer Posts
Show Posts	All Only With Images Only Without Images
Deleted Posts	All Only Deleted Posts Only Non-Deleted Posts
Ghost Posts	All Only Ghost Posts Only Non-Ghost Posts
Post Type	All Only Sticky Threads Only Opening Posts Only Reply Posts
Results	All Grouped By Threads
Order	Latest Posts First Oldest Posts First

Your latest searches