|
aku terangkan secara ringkas jelah ye
aku amik MSc.Quantitaive Sciences
tempoh setahun 1/2
course aku nie lebih kepada business&saham tp applied statistics
antara subjects yg kena amik:
d/ware house
forecasting
econometrics
protfolio theory
securities and market
stochastics
finance and risk analysis
and some business tpoics
emm...penat gak nak abihkan master...tp berbaloi...alhamdulillah
klu ko nak info lagi leh la buka website nih
www.tmsk.itm.edu.my |
|
|
|
|
|
|
ancoi This user has been deleted
|
EM nie macam method of moment ke? Adakah extension die GMM? Nie macam time series aje...:stp: |
|
|
|
|
|
|
|
EM ni method utk cari parameter bila ader missing values...so kita leh estimate dia nyer parameter. Em leh guna utk pelbagai jenis data and distribution...but itulah...setakat nie tgh try Rubin's Special formula utk EM nih... |
|
|
|
|
|
|
ancoi This user has been deleted
|
OOO.... aku paham, aku dah salah tajukle nie...ff: ingatkan nak bincang pasal time series nyer... yang aku tahu, ARCH n GARCH aje... |
|
|
|
|
|
|
|
ooo....itu utk time series...takpe2 leh gak...
aku nyer data ni...data time series gak tapi univariate...
skang ni tgh nak setelkan klu ader missing values...so nak replacekan data yg missing tu guna em...tgh pk bila yek nak dpt value tu...buat masa skang ni, yg leh terpk hanya dapat je estimation utk parameter then generate guna normal. tp tak sure lagi...itu yg tgh carik LR lagi ni....tapi tak jumpa2...so kalau ancoi ader dpt apa2 info bg tau yek....TQ |
|
|
|
|
|
|
|
hai;
leh tak sape2 kongsi ngan aku camner yek EM Algorithm tu impute missing data...emm...pening gak pk...
Aku dah wat guna Little&Rubin's special formula, leh tak sape2 tolong cek sama ada betul tak jln aku nie:
Initialization
1. aku estimate parameter cam mean and std deviation guna basic stat...tapi aku anik yg complete je...
e-step (complete data je)
2. aku guna special formula tu carikan estimate parameters (mean and std deviation)
m-step
3. based on special formula tu aku substitute apa yg dpt kat e-step tu...pastu aku iterate sampai std deviation converge....
lastly...aku generate data based on MLE yg last...then aku fill up missing value dgn data yg aku generate td...
skang ni aku tgh conpius...leh ke aku wat camnie...sbb dlm byk2 artikel yg aku baca takde pun yg citer pasal impute data camner...atau aku yg tak perasannn....hehehehhe....
tolong yek pada saper2 yg tau...
Thanks a lot...
:love: |
|
|
|
|
|
|
| |
Category: Belia & Informasi
|