P. 1
IMPLEMENTASI METODE FUZZY TIME SERIES DENGAN PENENTUAN INTERVAL BERBASIS RATA-RATA UNTUK PERAMALAN DATA PENJUALAN BULANAN

IMPLEMENTASI METODE FUZZY TIME SERIES DENGAN PENENTUAN INTERVAL BERBASIS RATA-RATA UNTUK PERAMALAN DATA PENJUALAN BULANAN

4.0

|Views: 3,327|Likes:
Published by Syauqi Haris
Fuzzy time series forecasting system seizes the patterns from the previous data and uses it to project the future data. Whereas it does not need a sophisticated learning method as in genetic algorithm and neural network does, will make fuzzy time series easier to be developed. In the forecasting calculation using this method, interval length is determined at the beginning of the process. While the different lengths of intervals may results vary in fuzzy relationship, and in turn different forecasting results. Therefore, the appropriate fuzzy relationship is important to be built based on the effective length of intervals. One of the solutions is average-based interval fuzzy time series. In this thesis, average-based fuzzy time series is implemented to forecast the amount of monthly sales that available on the U.S. Census Bureau website. From the research can be obtained that average-based interval fuzzy time series is more accurate than the standard method with fixed number of intervals. The calculation shows approximately 52.39 % in AFER and 70.90 % in MSE for the result deviation between the two methods.
Fuzzy time series forecasting system seizes the patterns from the previous data and uses it to project the future data. Whereas it does not need a sophisticated learning method as in genetic algorithm and neural network does, will make fuzzy time series easier to be developed. In the forecasting calculation using this method, interval length is determined at the beginning of the process. While the different lengths of intervals may results vary in fuzzy relationship, and in turn different forecasting results. Therefore, the appropriate fuzzy relationship is important to be built based on the effective length of intervals. One of the solutions is average-based interval fuzzy time series. In this thesis, average-based fuzzy time series is implemented to forecast the amount of monthly sales that available on the U.S. Census Bureau website. From the research can be obtained that average-based interval fuzzy time series is more accurate than the standard method with fixed number of intervals. The calculation shows approximately 52.39 % in AFER and 70.90 % in MSE for the result deviation between the two methods.

More info:

Categories:Types, Research
Published by: Syauqi Haris on Jun 13, 2010
Copyright:Traditional Copyright: All rights reserved

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF or read online from Scribd
See more
See less

11/28/2013

Himpunan fuzzy yang telah terbentuk kemudian ditampilkan
pada sebuah komponen TChart sebagaimana dijelaskan pada Kode
Program 4.11.

01
02
03
04
05

procedure TForm2.Button3Click(Sender: TObject);
var
i,j : integer;
seri : TLineSeries;
begin

52

06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35

with Chart1 do
begin

Chart1.FreeAllSeries;
Title.Text.Clear;
Title.Text.Add('Grafik Himpunan Fuzzy');
SetLength(nilai_tengah,m);
for i:=0 to m-1 do
begin

seri := TLineSeries.Create(nil);
seri.Clear;
seri.Title := 'fuzzyset';
seri.ParentChart := Chart1;
for j:=0 to 2 do
begin
if
(floor(interval_fuzzy[i,j])>=floor(terkecil))
and
(floor(interval_fuzzy[i,j])<=floor(terbesar)) then
begin

case j of

0:seri.AddXY(interval_fuzzy[i,j],0);
1:seri.AddXY(interval_fuzzy[i,j],1);
2:seri.AddXY(interval_fuzzy[i,j],0);

end;

end;

end;
nilai_tengah[i] := interval_fuzzy[i,1];

end;

end;

end;

Kode Program 4.11 Proses Menampilkan Fuzzy Sets dalam Grafik

Masing-masing data time series yang menjadi input dalam
program harus dihitung nilai keanggotaannya atas tiap-tiap himpunan
fuzzy yang terbentuk untuk kepentingan perhitungan selanjutnya
tentang fuzzifikasi. Hal ini dijelaskan pada Kode Program 4.12.

01
02
03
04
05
06
07
08
09
10

procedure TForm2.Button4Click(Sender: TObject);
var
i, j : Integer;
begin

StringGrid3.ColCount := m+1;
StringGrid3.RowCount := n+1;
SetLength(fuzzified,n+1,m+1);
StringGrid3.Refresh;
for i:=0 to n-1 do
begin

53

11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

for j:=0 to m-1 do
begin

if (dataTS[i] >= interval_fuzzy[j,0]) and
(dataTS[i] <= interval_fuzzy[j,2]) then
begin

fuzzified[i,j]:=((selisih)-Abs(dataTS[i]-
interval_fuzzy[j,1]))/(selisih);
fuzzified[i,j]:=RoundTo(fuzzified[i,j],-3);

end;

end;

end;
for j:=0 to m-1 do
StringGrid3.Cells[j+1,0]:=FloatToStr(interval_fuzzy
[j,1]);

for i:=0 to n-1 do
begin

StringGrid3.Cells[0,i+1]:=FloatToStr(dataTS[i]);
for j:=0 to m-1 do
begin
StringGrid3.Cells[j+1,i+1]:=FloatToStr(fuzzified[i,
j]);

end;

end;

end;

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->