The Student Room Group

My uphill battle with Statistics

I'm your typical engineering student that's done only mechanics modules before and while now at uni, this is my first experience with statistics.

I've been trying to get better but not hard enough so I'm making this thread to keep a public record of my progress.

I'll also be asking question (not for full solutions this time :biggrin:) but I'll post a question and may ask for a numerical answer to check my work to I'll post photos to ask if they're correct :smile:

And don't be shy other people, you can ask your statistics inquiries here as well, we can all learn together :smile:

Happy Posting! :colondollar::colondollar:

@Zacken (tag others that may be interested) :redface:
(edited 8 years ago)

Scroll to see replies

Remember.

AND rule is multiplication
OR rule is add
Remember.

Hakuna /=/ Matata
Reply 3
Sounds good, I'll be happy to help with what I can.
Reply 4
Man where was this help when i was failing 2 years ago lol
Reply 5
Original post by MevMev
Man where was this help when i was failing 2 years ago lol


Right here, the maths forum has been going on for a decade or so.
Original post by Zacken
Sounds good, I'll be happy to help with what I can.


bayes.jpg

For part a, I got 0.1 but they say the answer is 0.28?

I tried using the conditional probability P(IC | CP) = P(IC CP) / P(CP) = (0.7 x 0.1)/07 = 0.1

Note: IC - internal corrosion, CP - corrosion present , CF = corrosion falsely detected

My issues
1) Not entirely sure how to get the intersection part in the formula
2) I don't understand partitioning events with its compliment.
Original post by ServantOfMorgoth
bayes.jpg

For part a, I got 0.1 but they say the answer is 0.28?

I tried using the conditional probability P(IC | CP) = P(IC CP) / P(CP) = (0.7 x 0.1)/07 = 0.1

Note: IC - internal corrosion, CP - corrosion present , CF = corrosion falsely detected

My issues
1) Not entirely sure how to get the intersection part in the formula
2) I don't understand partitioning events with its compliment.


You're getting confused on what is conditioned on what. So let D be the event that the test detects corrosion and let CP be the event "corrosion present". Then you are given P(D | CP) = 0.7 and that P(D | ~CP) = 0.2 and P(CP) = 0.1.

You are asked to work out P(CP | D). Can you do that via Bayes' theorem? That is

P(CPD)=P(DCP)P(CP)P(DCP)P(CP)+P(DCP)P(CP)\displaystyle P(CP | D) = \frac{P(D | CP) P(CP)}{P(D | CP) P(CP) + P(D | \sim CP) P(\sim CP)}
(edited 8 years ago)
Original post by Gregorius
You're getting confused on what is conditioned on what. So let D be the event that the test detects corrosion and let CP be the event "corrosion present". Then you are given P(D | CP) = 0.7 and that P(D | ~CP) = 0.2 and P(CP) = 0.1.

You are asked to work out P(CP | D). Can you do that via Bayes' theorem? That is

P(CPD)=P(DCP)P(CP)P(DCP)P(CP)+P(DCP)P(CP)\displaystyle P(CP | D) = \frac{P(D | CP) P(CP)}{P(D | CP) P(CP) + P(D | \sim CP) P(\sim CP)}


What does the sum of the denominator symbolize? Like I only know Bayes theorem to be P(AjB)=P(Aj).P(BAj)i=1nP(Ai).P(BAi)P(A_j |B) = \dfrac{P(A_j) . P(B|A_j)}{\displaystyle\sum_{i=1}^n P(A_i).P(B|A_i)}
Original post by Zacken
Right here, the maths forum has been going on for a decade or so.


What if you post too much surely you'd get annoyed :colondollar:
Original post by ServantOfMorgoth
What does the sum of the denominator symbolize? Like I only know Bayes theorem to be P(AjB)=P(Aj).P(BAj)i=1nP(Ai).P(BAi)P(A_j |B) = \dfrac{P(A_j) . P(B|A_j)}{\displaystyle\sum_{i=1}^n P(A_i).P(B|A_i)}


The sum in the denominator I have given is a special case of the one you have there. You have A1,A2,AnA_1, A_2, \cdots A_n as a partition of the the probability space Ω\Omega; I have CP CP and not CPCP partitioning Ω\Omega.
Original post by Gregorius
The sum in the denominator I have given is a special case of the one you have there. You have A1,A2,AnA_1, A_2, \cdots A_n as a partition of the the probability space Ω\Omega; I have CP CP and not CPCP partitioning Ω\Omega.


Oh ok, I think I understand but in part b I think I need to find P(CP|~D) ? I'm not sure how to operate with ~D since they didn't give D separately?
Original post by Samistrawberry
What if you post too much surely you'd get annoyed :colondollar:


Surely they would over time but there's lots of people to help :smile:
Reply 13
Original post by Samistrawberry
What if you post too much surely you'd get annoyed :colondollar:


That would defeat the purpose of the forum, everybody is welcome to post as many question as they'd like. I've helped a fair few people on here, and amongst those who post repeated questions, I can never once say that I felt annoyed by them as long as they were courteous.
Original post by ServantOfMorgoth
Oh ok, I think I understand but in part b I think I need to find P(CP|~D) ? I'm not sure how to operate with ~D since they didn't give D separately?


What you can work out from what they have given you is P(~D | CP) = 1 - P(D | CP). You can now use Bayes to invert that one.
Original post by ServantOfMorgoth
Surely they would over time but there's lots of people to help :smile:


Haha I've stopped studying maths but they're a lot more fun to help solve then chem questions.

Original post by Zacken
That would defeat the purpose of the forum, everybody is welcome to post as many question as they'd like. I've helped a fair few people on here, and amongst those who post repeated questions, I can never once say that I felt annoyed by them as long as they were courteous.


Sometimes I'm just too dumb to understand and it might frustrate others.
Reply 16
Original post by Samistrawberry

Sometimes I'm just too dumb to understand and it might frustrate others.


Give it a try next time you get stuck, you'll be pleasantly surprised.
I'll be back on Wednesday afternoon as per my schedule permits :biggrin:

Thanks all so far @Gregorius
Original post by Zacken
Give it a try next time you get stuck, you'll be pleasantly surprised.


Chemistry is a slight bit more confusing to understand/explain. I often just shake my head in class saying I understand when I don't when miss explains.

Will try to make use of the Chem forum and try not to spend all my time on the Chat section :teehee:
(edited 8 years ago)
Guys I flopped my statistics incourse test :sad:

Quick Reply

Latest