Tugas Kelompok - Metode Peramalan Deret Waktu
P1 - Kelompok 7 STA1341 2022
Anggota Kelompok:
- Aprilia Permata Putri (G1401201002)
- Nana Oktaviana (G1401201006)
- Ferista Wahyu Saputri (G1401201008)
- Ainaini Salsabila (G1401201055)
- Akmal Riza Wibisono (G1401201086)
PEMULUSAN DATA TIME SERIES
Pra-Processing Data
Memanggil Package di R
Untuk melakukan proses pengecekan kestasioneran data, diperlukan beberapa library untuk menunjang proses analisis data sebagai berikut.
library(googlesheets4)
library(tseries)
library(forecast)
library(TTR)
library(TSA)
library(imputeTS)
library(forecast)
library(graphics)
library(dplyr)
require(smooth)
require(Mcomp)
library(fpp2)Import Data
Data yang digunakan hanyalah data Wisatawan dengan syntax berikut:
gs4_deauth()
wisata <- read_sheet("https://docs.google.com/spreadsheets/d/1oWrxKrK3FVRTbVCGb-ijLp0OY8_WH5dChhS4CvZnk6k/edit?usp=sharing")
wisata <- wisata[c("Periode Waktu","Jumlah Kunjungan")]
wisata## # A tibble: 114 x 2
## `Periode Waktu` `Jumlah Kunjungan`
## <dttm> <dbl>
## 1 2013-01-01 00:00:00 614328
## 2 2013-02-01 00:00:00 678415
## 3 2013-03-01 00:00:00 725316
## 4 2013-04-01 00:00:00 646117
## 5 2013-05-01 00:00:00 700708
## 6 2013-06-01 00:00:00 789594
## 7 2013-07-01 00:00:00 717784
## 8 2013-08-01 00:00:00 771009
## 9 2013-09-01 00:00:00 770878
## 10 2013-10-01 00:00:00 719903
## # ... with 104 more rowsAnalisis Eksplorasi Data
plot_wisata = ts(wisata$`Jumlah Kunjungan`, frequency = 12, start= 2013)
plot(plot_wisata, main = "Kunjungan Wisatawan Mancanegara Periode 2013-2022",
xlab = "Tahun", ylab="Wisatawan")
points(plot_wisata,col="red")
Data historis yang digunakan merupakan data sekunder sejumlah 114 amatan yang bersumber dari situs resmi Badan Pusat Statistik (bps.go.id) berupa total kunjungan bulanan wisatawan mancanegara ke Indonesia dari periode Januari 2013 hingga Juni 2022.
Berdasarkan grafik deret waktu, dapat diamati bahwa total kunjungan bulanan wisatawan mancanegara ke Indonesia terus mengalami peningkatan sejak tahun 2013 hingga awal tahun 2020. Hal ini menjadi salah satu indikator keberhasilan pemerintah Indonesia, khususnya Kementerian Pariwisata, dalam mempromosikan destinasi wisata nasional di lingkup internasional.
Terhitung sejak pandemi COVID-19 melanda dunia sejak Maret 2020, total kunjungan wisatawan mancanegara ke Indonesia ikut menurun drastis 88% hingga ke titik 150.000 kunjungan wisatawan per bulan. Sejak April 2020 hingga saat ini, total kunjungan wisatawan masih stabil dengan rata-rata 150.000 wisatawan per bulannya meskipun sudah mulai menunjukkan tren peningkatan di pertengahan tahun 2022 seiring dengan mulai meredanya penyebaran virus korona di tengah masyarakat internasional.
Split Data Training dan Testing (90% dan 10%)
train <- wisata[1:102,2]
test <- wisata[103:114,2]Data Time Series
Membentuk objek time series
winter.ts <- ts(wisata$`Jumlah Kunjungan`, start=2013, frequency = 12)
training.ts <- ts(train, start=2013, frequency = 12)
testing.ts <- ts(test, start=2021, frequency = 12)Membuat Plot Time Series
plot(winter.ts, col="red", main = "Kunjungan Wisatawan Mancanegara Periode 2013-2022", xlab = "Tahun", ylab="Wisatawan")
points(winter.ts)
plot(training.ts, col="blue", main = "Kunjungan Wisatawan Mancanegara Periode 2013-2021", xlab = "Tahun", ylab="Wisatawan")
points(training.ts)
Metode Smoothing
Double Exponential Smoothing
df_des1 <- HoltWinters(training.ts, alpha = 0.9, beta=0.1, gamma=F)
df_des1## Holt-Winters exponential smoothing with trend and without seasonal component.
##
## Call:
## HoltWinters(x = training.ts, alpha = 0.9, beta = 0.1, gamma = F)
##
## Smoothing parameters:
## alpha: 0.9
## beta : 0.1
## gamma: FALSE
##
## Coefficients:
## [,1]
## a 124733.05
## b -25476.52df_des1$fitted## xhat level trend
## Mar 2013 742502.00 678415.0 64087.0000
## Apr 2013 789574.86 727034.6 62540.2600
## May 2013 710091.84 660462.8 49629.0526
## Jun 2013 750430.89 701646.4 48784.5071
## Jul 2013 837986.88 785677.7 52309.1869
## Aug 2013 771295.22 729804.3 41490.9281
## Sep 2013 812502.79 771037.6 41465.1687
## Oct 2013 812759.42 775040.5 37718.9376
## Nov 2013 758550.50 729188.6 29361.8601
## Dec 2013 836295.15 802534.9 33760.2949
## Jan 2014 894171.70 858219.0 35952.6819
## Feb 2014 790442.61 767188.3 23254.3392
## Mar 2014 726798.11 711443.7 15354.4444
## Apr 2014 780573.36 761726.1 18847.2449
## May 2014 745721.66 731756.1 13965.5229
## Jun 2014 766262.11 751698.9 14563.2437
## Jul 2014 865186.11 842953.7 22232.4038
## Aug 2014 800322.16 786007.6 14314.5535
## Sep 2014 840870.57 824171.1 16699.4486
## Oct 2014 808491.19 796253.5 12237.7378
## Nov 2014 821001.98 808739.4 12262.5603
## Dec 2014 777288.97 770115.1 7173.8721
## Jan 2015 921127.42 901529.5 19597.9248
## Feb 2015 746260.22 744340.9 1919.2768
## Mar 2015 795740.86 789497.8 6243.0371
## Apr 2015 799076.41 793097.7 5978.7198
## May 2015 757458.49 755806.7 1651.7533
## Jun 2015 795577.40 790610.4 4966.9489
## Jul 2015 820076.65 813334.0 6742.6130
## Aug 2015 822140.87 815823.6 6317.3043
## Sep 2015 859250.27 850133.7 9116.5860
## Oct 2015 879356.58 869240.9 10115.6514
## Nov 2015 836843.26 831512.1 5331.1993
## Dec 2015 783895.87 783862.7 33.1462
## Jan 2016 912561.82 900834.8 11727.0377
## Feb 2016 827012.63 824128.9 2883.7435
## Mar 2016 890579.78 882179.4 8400.4171
## Apr 2016 923175.02 912575.1 10599.9470
## May 2016 911915.75 903303.0 8612.7447
## Jun 2016 923785.84 914877.0 8908.8675
## Jul 2016 867221.22 864264.5 2956.7317
## Aug 2016 1034042.53 1016189.0 17853.5122
## Sep 2016 1049860.08 1032191.7 17668.4242
## Oct 2016 1024753.49 1010973.7 13779.7872
## Nov 2016 1054271.81 1039061.2 15210.5627
## Dec 2016 1018062.95 1007526.9 10536.0696
## Jan 2017 1122911.42 1103801.5 19109.9240
## Feb 2017 1127227.36 1109462.3 17765.0163
## Mar 2017 1042191.41 1033771.9 8419.4740
## Apr 2017 1068020.62 1058018.4 10002.1772
## May 2017 1180354.52 1161049.5 19305.0615
## Jun 2017 1168210.73 1151764.7 16446.0744
## Jul 2017 1160689.17 1146422.0 14267.1990
## Aug 2017 1382759.18 1349600.8 33158.3636
## Sep 2017 1426296.53 1392194.6 34101.9073
## Oct 2017 1286093.56 1267837.6 18256.0100
## Nov 2017 1181066.30 1174017.9 7048.4394
## Dec 2017 1070268.80 1073933.6 -3664.8272
## Jan 2018 1142598.55 1139354.8 3243.7706
## Feb 2018 1101530.37 1102315.0 -784.5890
## Mar 2018 1195758.68 1187905.7 7852.9480
## Apr 2018 1369602.27 1346659.3 22943.0064
## May 2018 1325936.82 1309049.1 16887.6917
## Jun 2018 1260425.01 1251028.2 9396.8280
## Jul 2018 1331448.34 1316449.1 14999.2371
## Aug 2018 1560072.41 1525652.7 34419.6766
## Sep 2018 1545931.19 1515926.1 30005.0497
## Oct 2018 1402697.93 1388441.8 14256.1125
## Nov 2018 1306972.04 1302714.3 4257.7487
## Dec 2018 1163235.64 1172431.9 -9196.2651
## Jan 2019 1393934.55 1381322.2 12612.3874
## Feb 2019 1216269.38 1220955.0 -4685.5722
## Mar 2019 1239033.16 1241223.3 -2190.1767
## Apr 2019 1308992.04 1304623.2 4368.8288
## May 2019 1278947.44 1277707.1 1240.3347
## Jun 2019 1251070.45 1252477.1 -1406.6948
## Jul 2019 1430865.98 1415799.7 15066.2348
## Aug 2019 1482866.16 1464442.3 18423.8666
## Sep 2019 1548217.85 1525527.8 22690.0318
## Oct 2019 1413004.02 1404668.9 8335.1355
## Nov 2019 1355434.84 1353091.0 2343.8336
## Dec 2019 1283871.37 1288246.4 -4375.0116
## Jan 2020 1371760.03 1367747.4 4012.5949
## Feb 2020 1295237.09 1298545.9 -3308.8180
## Mar 2020 873680.90 915012.2 -41331.3056
## Apr 2020 448698.95 524907.6 -76208.6369
## May 2020 84763.69 187129.3 -102365.6027
## Jun 2020 58705.61 154134.2 -95428.5550
## Jul 2020 60153.89 146775.5 -86621.5703
## Aug 2020 68164.55 146183.2 -78018.6405
## Sep 2020 82596.51 152210.6 -69614.0399
## Oct 2020 78706.09 142345.3 -63639.1662
## Nov 2020 87917.96 144934.3 -57016.3439
## Dec 2020 86894.08 138820.2 -51926.1207
## Jan 2021 111381.03 156360.5 -44979.4775
## Feb 2021 81384.18 125001.6 -43617.4202
## Mar 2021 61926.54 103347.6 -41421.0767
## Apr 2021 77977.40 114173.8 -36196.3554
## May 2021 76211.86 109278.1 -33066.2813
## Jun 2021 105734.51 133110.9 -27376.3786Akurasi Data Training
sse1 <- df_des1$SSE
mse1 <- sse1/length(training.ts)
rmse1 <-sqrt(mse1)
akurasi1 <- c("SSE"=sse1, "MSE"=mse1, "RMSE"=rmse1)
akurasi1## SSE MSE RMSE
## 1.222513e+12 1.198542e+10 1.094780e+05Pemulusan DES dengan Alpha Beta Optimum
# ITERASI MENCARI NILAI ALPHA BETA OPTIMUM
a = c(0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9)
b = c(0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9)
output1 = data.frame()
for (i in a) {
for (j in b) {
des1 <- HoltWinters(training.ts, alpha = i, beta=j, gamma=F)
sse1 <- des1$SSE
mse1 <- sse1/length(training.ts)
rmse1 <-sqrt(mse1)
akurasi1 <- cbind("SSE"=sse1, "MSE"=mse1, "RMSE"=rmse1)
output1 <- rbind(output1, akurasi1)
}
}
output_des1 <- cbind("Alpha" = rep(c(0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9), each=9), "Beta"=b, output1)
output_des1## Alpha Beta SSE MSE RMSE
## 1 0.1 0.1 8.300512e+12 81377569954 285267.5
## 2 0.1 0.2 6.319724e+12 61958074637 248913.8
## 3 0.1 0.3 6.305676e+12 61820353376 248637.0
## 4 0.1 0.4 6.536981e+12 64088046458 253156.2
## 5 0.1 0.5 6.719991e+12 65882268895 256675.4
## 6 0.1 0.6 6.717820e+12 65860981222 256633.9
## 7 0.1 0.7 6.578284e+12 64492981631 253954.7
## 8 0.1 0.8 6.510730e+12 63830685653 252647.4
## 9 0.1 0.9 6.580969e+12 64519304985 254006.5
## 10 0.2 0.1 4.092829e+12 40125771374 200314.2
## 11 0.2 0.2 4.017021e+12 39382559860 198450.4
## 12 0.2 0.3 4.063170e+12 39834998769 199587.1
## 13 0.2 0.4 4.077612e+12 39976589941 199941.5
## 14 0.2 0.5 4.149743e+12 40683759708 201702.2
## 15 0.2 0.6 4.278384e+12 41944944123 204804.6
## 16 0.2 0.7 4.394697e+12 43085267604 207569.9
## 17 0.2 0.8 4.425173e+12 43384049494 208288.4
## 18 0.2 0.9 4.341777e+12 42566442488 206316.4
## 19 0.3 0.1 2.851895e+12 27959753984 167211.7
## 20 0.3 0.2 2.835231e+12 27796382308 166722.5
## 21 0.3 0.3 2.844583e+12 27888065671 166997.2
## 22 0.3 0.4 2.865362e+12 28091779773 167606.0
## 23 0.3 0.5 2.865492e+12 28093062176 167609.9
## 24 0.3 0.6 2.814049e+12 27588719575 166098.5
## 25 0.3 0.7 2.721521e+12 26681580114 163345.0
## 26 0.3 0.8 2.619594e+12 25682297884 160257.0
## 27 0.3 0.9 2.533766e+12 24840846754 157609.8
## 28 0.4 0.1 2.193549e+12 21505384775 146647.1
## 29 0.4 0.2 2.175004e+12 21323572742 146025.9
## 30 0.4 0.3 2.167874e+12 21253670792 145786.4
## 31 0.4 0.4 2.149419e+12 21072734192 145164.5
## 32 0.4 0.5 2.110741e+12 20693538220 143852.5
## 33 0.4 0.6 2.064370e+12 20238921741 142263.6
## 34 0.4 0.7 2.026774e+12 19870336950 140962.2
## 35 0.4 0.8 2.005558e+12 19662328547 140222.4
## 36 0.4 0.9 2.001049e+12 19618125003 140064.7
## 37 0.5 0.1 1.801492e+12 17661684735 132897.3
## 38 0.5 0.2 1.785219e+12 17502149347 132295.7
## 39 0.5 0.3 1.774745e+12 17399459880 131907.0
## 40 0.5 0.4 1.758997e+12 17245070049 131320.5
## 41 0.5 0.5 1.742804e+12 17086316433 130714.6
## 42 0.5 0.6 1.735423e+12 17013951193 130437.5
## 43 0.5 0.7 1.741618e+12 17074681975 130670.1
## 44 0.5 0.8 1.761950e+12 17274023421 131430.7
## 45 0.5 0.9 1.795307e+12 17601046704 132668.9
## 46 0.6 0.1 1.554628e+12 15241449877 123456.3
## 47 0.6 0.2 1.546069e+12 15157544040 123116.0
## 48 0.6 0.3 1.544217e+12 15139381174 123042.2
## 49 0.6 0.4 1.544828e+12 15145373039 123066.5
## 50 0.6 0.5 1.552557e+12 15221143610 123374.0
## 51 0.6 0.6 1.571147e+12 15403405034 124110.5
## 52 0.6 0.7 1.601014e+12 15696211208 125284.5
## 53 0.6 0.8 1.640429e+12 16082636915 126817.3
## 54 0.6 0.9 1.686596e+12 16535253631 128589.5
## 55 0.7 0.1 1.393924e+12 13665919193 116901.3
## 56 0.7 0.2 1.395519e+12 13681558512 116968.2
## 57 0.7 0.3 1.405462e+12 13779036681 117384.1
## 58 0.7 0.4 1.421523e+12 13936497306 118052.9
## 59 0.7 0.5 1.446015e+12 14176621583 119065.6
## 60 0.7 0.6 1.479508e+12 14504983451 120436.6
## 61 0.7 0.7 1.520456e+12 14906434806 122091.9
## 62 0.7 0.8 1.566320e+12 15356082103 123919.7
## 63 0.7 0.9 1.614529e+12 15828717332 125812.2
## 64 0.8 0.1 1.288592e+12 12633252108 112397.7
## 65 0.8 0.2 1.300991e+12 12754811242 112937.2
## 66 0.8 0.3 1.322694e+12 12967584331 113875.3
## 67 0.8 0.4 1.351755e+12 13252504434 115119.5
## 68 0.8 0.5 1.388789e+12 13615580462 116685.8
## 69 0.8 0.6 1.433095e+12 14049952640 118532.5
## 70 0.8 0.7 1.483071e+12 14539916369 120581.6
## 71 0.8 0.8 1.537245e+12 15071028426 122764.1
## 72 0.8 0.9 1.594963e+12 15636893657 125047.6
## 73 0.9 0.1 1.222513e+12 11985424883 109478.0
## 74 0.9 0.2 1.246380e+12 12219414836 110541.5
## 75 0.9 0.3 1.280312e+12 12552074479 112036.0
## 76 0.9 0.4 1.322493e+12 12965613461 113866.6
## 77 0.9 0.5 1.372996e+12 13460748488 116020.5
## 78 0.9 0.6 1.431368e+12 14033022590 118461.1
## 79 0.9 0.7 1.497216e+12 14678591984 121155.2
## 80 0.9 0.8 1.570910e+12 15401079268 124101.1
## 81 0.9 0.9 1.653861e+12 16214325839 127335.5datades1 <- data.frame(training.ts, c(NA, NA, df_des1$fitted[,1]))
colnames(datades1) = c("y","yhat")
head(datades1)## y yhat
## 1 614328 NA
## 2 678415 NA
## 3 725316 742502.0
## 4 646117 789574.9
## 5 700708 710091.8
## 6 789594 750430.9Plot Hasil Pemulusan DES
ts.plot(training.ts,xlab="Periode Waktu", ylab="Jumlah Wisatawan", col="blue",
lty=2, ylim=c(0,1600000))
points(training.ts)
lines (datades1[,2], col="red",lwd=2)
title("Double Exponential Smoothing (Alpha=0.9 dan Beta 0.1)", cex.main=1, font.main=4 ,col.main="black")
legend("topleft", c("Data aktual","Fitted DES"), lty=1:3,col=c ("blue","red"))
Hasil Peramalan
ramal_des1 <- forecast::forecast(df_des1,h=12)
(df_ramal_des1 <- data.frame(ramal_des1))## Point.Forecast Lo.80 Hi.80 Lo.95 Hi.95
## Jul 2021 99256.53 -42577.06 241090.1 -117659.2 316172.3
## Aug 2021 73780.00 -125802.58 273362.6 -231455.2 379015.2
## Sep 2021 48303.48 -203286.27 299893.2 -336469.8 433076.8
## Oct 2021 22826.95 -278561.90 324215.8 -438107.5 483761.4
## Nov 2021 -2649.57 -353038.81 347739.7 -538523.7 533224.5
## Dec 2021 -28126.09 -427419.75 371167.6 -638793.0 582540.9
## Jan 2022 -53602.62 -502099.61 394894.4 -739519.6 632314.3
## Feb 2022 -79079.14 -577316.93 419158.6 -841068.1 682909.8
## Mar 2022 -104555.67 -653222.53 444111.2 -943669.2 734557.9
## Apr 2022 -130032.19 -729914.48 469850.1 -1047473.0 787408.6
## May 2022 -155508.72 -807457.40 496440.0 -1152578.2 841560.8
## Jun 2022 -180985.24 -885893.87 523923.4 -1259050.0 897079.5Plot Peramalan
plot(ramal_des1, xlab="Periode Waktu", ylab="Jumlah Wisatawan")
Gabungan Data Aktual, Pemulusan, dan Ramalan
data.des1 <- cbind(aktual=c(training.ts, rep(NA,12)),
pemulusan=c(NA, df_des1$fitted[,2], as.numeric(df_des1$coefficients[1]+df_des1$coefficients[2]), rep(NA,12)),
ramalan = c(NA, NA, df_des1$fitted[,1], df_ramal_des1$Point.Forecast))
data.des1 <- ts(data.des1)Winter’s Aditif
Pemulusan
aditif <- HoltWinters(training.ts, seasonal = "additive")
aditif ## Holt-Winters exponential smoothing with trend and additive seasonal component.
##
## Call:
## HoltWinters(x = training.ts, seasonal = "additive")
##
## Smoothing parameters:
## alpha: 1
## beta : 0
## gamma: 0
##
## Coefficients:
## [,1]
## a 60940.979
## b 4278.619
## s1 -23074.063
## s2 23359.188
## s3 20538.937
## s4 -35457.146
## s5 46567.271
## s6 95069.604
## s7 -17560.854
## s8 -72775.437
## s9 -13010.687
## s10 -56839.104
## s11 -32720.729
## s12 65903.021Forecasting
ramalan1 <- forecast(aditif, h=12)
ramalan1## Point Forecast Lo 80 Hi 80 Lo 95 Hi 95
## Jul 2021 42145.54 -109977.0 194268.1 -190505.8 274796.8
## Aug 2021 92857.40 -122276.3 307991.1 -236161.2 421876.0
## Sep 2021 94315.77 -169168.2 357799.7 -308648.1 497279.7
## Oct 2021 42598.31 -261646.7 346843.3 -422704.3 507900.9
## Nov 2021 128901.34 -211254.9 469057.6 -391322.8 649125.5
## Dec 2021 181682.29 -190940.3 554304.8 -388194.7 751559.3
## Jan 2022 73330.46 -329147.9 475808.8 -542207.1 688868.0
## Feb 2022 22394.49 -407873.0 452661.9 -635642.8 680431.8
## Mar 2022 86437.86 -369929.7 542805.4 -611516.1 784391.8
## Apr 2022 46888.06 -434165.6 527941.7 -688820.0 782596.1
## May 2022 75285.05 -429248.3 579818.4 -696332.1 846902.2
## Jun 2022 178187.42 -348780.4 705155.3 -627740.4 984115.2Akurasi data training
sse1.train <- aditif$SSE
sse1.train## [1] 1.266537e+12Plot Pemulusan
plot(ramalan1, xlab="Periode Waktu", ylab="Jumlah Kunjungan")
Winter’s Multiplikatif
Pemulusan
multi <- HoltWinters(training.ts, seasonal = "multiplicative")
multi ## Holt-Winters exponential smoothing with trend and multiplicative seasonal component.
##
## Call:
## HoltWinters(x = training.ts, seasonal = "multiplicative")
##
## Smoothing parameters:
## alpha: 0.8255676
## beta : 0.05675182
## gamma: 1
##
## Coefficients:
## [,1]
## a 1.190900e+05
## b -2.813304e+04
## s1 1.132100e+00
## s2 1.128662e+00
## s3 1.043335e+00
## s4 1.023979e+00
## s5 9.923898e-01
## s6 1.161302e+00
## s7 1.039876e+00
## s8 9.551069e-01
## s9 9.323945e-01
## s10 7.416652e-01
## s11 9.646528e-01
## s12 1.065110e+00Forecasting
ramalan2 <- forecast(multi, h=12)
ramalan2## Point Forecast Lo 80 Hi 80 Lo 95 Hi 95
## Jul 2021 102972.424 -35327.71 241272.6 -108539.4 314484.2
## Aug 2021 70907.004 -112386.36 254200.4 -209416.0 351230.0
## Sep 2021 36194.254 -176631.86 249020.4 -289295.2 361683.7
## Oct 2021 6715.131 -240387.20 253817.5 -371195.3 384625.5
## Nov 2021 -21410.970 -297648.93 254827.0 -443880.5 401058.5
## Dec 2021 -57726.228 -411781.22 296328.8 -599206.6 483754.2
## Jan 2022 -80945.259 -433729.40 271838.9 -620482.0 458591.5
## Feb 2022 -101216.762 -460170.41 257736.9 -650189.0 447755.5
## Mar 2022 -125040.926 -508987.01 258905.2 -712235.8 462153.9
## Apr 2022 -120328.020 -462368.90 221712.9 -643434.4 402778.4
## May 2022 -183644.218 -657200.15 289911.7 -907885.5 540597.1
## Jun 2022 -232733.412 -771386.99 305920.2 -1056533.0 591066.1Akurasi Data Training
sse2.train <- multi$SSE
sse2.train## [1] 1.043219e+12Plot Peramalan
plot(ramalan2, xlab="Periode Waktu", ylab="Jumlah Kunjungan")
Perbandingan Keakurasian Data Training
akurasi <- matrix(c(sse1.train, sse2.train), nrow=1, ncol=2)
row.names(akurasi)<- "SSE"
colnames(akurasi) <- c("Aditif", "Multiplikatif")
akurasi## Aditif Multiplikatif
## SSE 1.266537e+12 1.043219e+12Keakurasian Data Testing Metode Winter’s Multiplikatif
selisih1<-as.numeric(ramalan1$mean)-as.numeric(testing.ts)
selisih1## [1] -85103.46 -25675.60 -25784.23 -103538.69 -21675.66 18063.29
## [7] -70247.54 3939.49 45647.86 -64168.94 -137046.95 -167250.58SSEtesting1<-sum(selisih1^2)
selisih2<-as.numeric(ramalan2$mean)-as.numeric(testing.ts)
selisih2## [1] -24276.58 -47626.00 -83905.75 -139421.87 -171987.97 -221345.23
## [7] -224523.26 -119671.76 -165830.93 -231385.02 -395976.22 -578171.41SSEtesting2<-sum(selisih2^2)Perbandingan Keakurasian Data Testing Metode Winter’s Aditif dan Multiplikatif
akurasi <- matrix(c(SSEtesting1, SSEtesting2), nrow=1, ncol=2)
row.names(akurasi)<- "SSE"
colnames(akurasi) <- c("Aditif", "Multiplikatif")
akurasi## Aditif Multiplikatif
## SSE 77989277550 744760086355metode yang dipilih : Multiplikatif