MedVision ad

bba315 (1 Viewer)

*hopeful*

Active Member
Joined
Oct 5, 2003
Messages
2,777
Location
earth
Gender
Female
HSC
2003
like how random is the subject, anyone else doing it ?

ooh i know skittled is because i saw him in my 'tute' :wave:
 
Joined
Sep 22, 2003
Messages
412
Gender
Female
HSC
2003
forecasting?

i hate it. hate it hate it hate it. will need to get a tutor. soon.

can't do the mathy subjects for shite.
 
Joined
Sep 22, 2003
Messages
412
Gender
Female
HSC
2003
on that note do u guys know how u go about getting a tutor? i think it's through efs but i forget
 

*hopeful*

Active Member
Joined
Oct 5, 2003
Messages
2,777
Location
earth
Gender
Female
HSC
2003
yeh scarecrow try thru efs/ecfs
very very shit subject
im still on qstn 1 for the workbook, how d hell r we supposed 2 do half the things, its not like we've been taught how to :|
 
Joined
Sep 22, 2003
Messages
412
Gender
Female
HSC
2003
i know it's fair tricky hey. i think we've just got to do the questions directly under the 'workbook collection' bit, so it's just look up some numbers, forecast ur uni mark/travel time and use a holts smoothing method. did u just use a naive forecast for the first one?

hey are u guys around tommorrow? want to work on our forecasting 'abilities' in the labs? try and get the assignment dispensed with?
 

*hopeful*

Active Member
Joined
Oct 5, 2003
Messages
2,777
Location
earth
Gender
Female
HSC
2003
haha im in the labs right now, being very secretive about being on bos
i wish it was as easy as u make it sound scarecrow!

naive forecast for the 1st one ? do u mean the travel time one ? we didnt use anything for that

hey how do u do the 'next 5 periods forecast' for question c part b ??? i hate this subject :(
 
Joined
Sep 22, 2003
Messages
412
Gender
Female
HSC
2003
i'm not sure... i haven't really tried to do the last questions yet i was planning on trying to crack it tommorrow not too cofident in my abilities there though - don't really understand Holts :(
 

Skittled

What did the crab do?
Joined
Jan 12, 2005
Messages
991
Location
Sydney, Australia
Gender
Male
HSC
2003
First off (becuase, everything below the quote is just a rant); Anyone up for a study group, closer to the end of sem?

*hopeful* said:
like how random is the subject, anyone else doing it ?

ooh i know skittled is because i saw him in my 'tute' :wave:
What is with BOS people, seeing other BOS people and not saying hello till you're online!? :burn:

lol. I don't mind the subject. Sorta approaching it with this "this is a copout" approach, because he reckon's (or implied, or something) regression is too complex to teach (given that I'm currently doing a unit on regression, I must disagree).

I also approach it with a "this is stupid", when they're testing our look-it-up-on-the-web skills (ie find the data for workbook.. question 4, or something). Also, the shitty tute instructions contribute to this impression.

So. Yeah. That's my opinion. Anyone got the text yet? I should get it; gonna go 2nd hand (anyone selling?), but just can't be bothered yet.

*hopeful* said:
hey how do u do the 'next 5 periods forecast' for question c part b ??? i hate this subject :(
I asked about that, from our tutor.. Apparently they'll "all be the same". I think that was his roundabout way of saying you can't, but the best you can do is a naive forecast for periods t+2 to t+5 (ie you can make a SES, or whatever it is, prediction for period 1, but period 2 will be a niave forecast, being t2=t1. t3 will =t2=t1, etc... Does that make sense?)
 
Joined
Sep 22, 2003
Messages
412
Gender
Female
HSC
2003
hey skittled!! definately up for a study group some time during sometime (as long as it doesn't bother you that i'm totally in love with your gf :))

i agree that this subject is.... interestingly taught - don't get me wrong i reckon con's a champ but the tute's are so big it makes it quite hard if you're struggling to get help

is the question u guys are stuck on ses? cause if so it would make sense if some of the forecasts are the same (since with ses i think u can only forecast like one period ahead if u don't have data for any of the other periods that follow). is holts the same kind of thing? damn holts! damn him!
 

sukiyaki

emptiness
Joined
Nov 9, 2002
Messages
1,505
Location
westie
Gender
Female
HSC
2003
sorry for the bump (and seems like the double post as well)

my friend brought this up and i didnt notice, but for example in question C we only do 3 right ? not 4 and 5 too, lol this whole time i though i had to do all 3 questions under the headling of questionn C and my friend told me its only question 3 itself for question C
 

Skittled

What did the crab do?
Joined
Jan 12, 2005
Messages
991
Location
Sydney, Australia
Gender
Male
HSC
2003
That's the way I interpreted it, yep

Qustion D:
WTF is the 'optimised' holts model vs the linear trend extrapolation? I didn't sit in the 2nd hour of week 4, which I assume is when he covered this (and everything re holts)...
 
Last edited:
Joined
Sep 22, 2003
Messages
412
Gender
Female
HSC
2003
sorry i'm not up to that yet... not really sure... at all

i'm still trying to find dow jones data to do the third question - i found it on yahoo finance but the data goes from 2007-2004 in order (not the other way around) so i can't use it for graphing or predicting which is not good. would anyone pretty please tell me either wherei can find it in the right order or how i can invert the order in the column so i can use it?? if i don't work it out soon i'll have to skip my mum's 60th birthday celebration (and become officially the black sheep in my family for a good long while)
 

*hopeful*

Active Member
Joined
Oct 5, 2003
Messages
2,777
Location
earth
Gender
Female
HSC
2003
yeah we (my friends n me) r also confused about wether to do 1 question per section or all under the underlined headings, talk about shit organisation

skittled: i will never say hi, i dont want 2 reveal myself :p
u do the holts thing and optimise the values of alpha and beta by using solver

scarecrow: play around with the 'sort' function in excel to invert the data

this assignment is a piece of shit and im probably gonna fail....only because i got no idea how to present my answers

piece of shit subject

/endrant
 

*hopeful*

Active Member
Joined
Oct 5, 2003
Messages
2,777
Location
earth
Gender
Female
HSC
2003
Skittled said:
I asked about that, from our tutor.. Apparently they'll "all be the same". I think that was his roundabout way of saying you can't, but the best you can do is a naive forecast for periods t+2 to t+5 (ie you can make a SES, or whatever it is, prediction for period 1, but period 2 will be a niave forecast, being t2=t1. t3 will =t2=t1, etc... Does that make sense?)
so we forecast t1 but t2,t3,t4,t5 also=t1 ?
 

swisstruffles

Member
Joined
Apr 1, 2003
Messages
169
Location
Sydney
Gender
Female
HSC
2003
i agree this subject is looking like a fail so far for me
those tutes r useless. the one tutor ratio against all the students is wayy too many. we never even get to go through the whole tute questions in time.
 
Joined
Sep 22, 2003
Messages
412
Gender
Female
HSC
2003
thanks hopeful i managed to invert it (proud moment)

i'm going to go ahead and assume it's just one question per section (though for d u've got to do the one above as well to find the answer to b)... u'd think they'd say 'questions' if they wanted us to do all the ones underneath. not very clear is it?

how r u guys presenting it? i mean there's like 9 pages of data for each of the C bits if u want to show all the working - do u think they really want to see that or should be chop some out for brevity?
 

*hopeful*

Active Member
Joined
Oct 5, 2003
Messages
2,777
Location
earth
Gender
Female
HSC
2003
i think we just have to show results, final pages etc

i dunno, im at the point where i dont really care anymore
 

Skittled

What did the crab do?
Joined
Jan 12, 2005
Messages
991
Location
Sydney, Australia
Gender
Male
HSC
2003
I'm so over this. This is the first assignment I've ever (including high school) handed in deliberately incomplete (leaving out Q4). Seriously can't be bothered.. :rolleyes:

I'm just giving interpretations. That is, graphs of the data collected, comments on the graphs, (level+trend, seasonality, cyclical factors, randomness).
For c, given specific stats (ie forecasts, error ratings), and comments about why this is better or that's better, etc. So, it's the final interprettion.

I have no intention on printing out any spreadsheets: how can you ever know what to print out? You've gotta have the data to do the above, so I think that'll do.

Save the trees! :D
 

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top