Perbandingan Keakuratan Hasil Peramalan Produksi Bawang Merah Metode Holt-Winters dengan Singular Spectrum Analysis (SSA)

The Holt-Winters method is used to model data with seasonal patterns, whether trends or not. There are two methods of forecasting in Singular Spectrum Analysis (SSA), namely recurrent method (R-forecasting) and vector method (V-forecasting). The recurrent method performs continuous continuation (with the help of LRF), whereas the vector method corresponds to the L-continuation. Different methods of course make a difference in the accuracy of forecast results. To see the difference between the three methods is done by looking at the comparison of accuracy and reliability of forecast results. To measure the accuracy of forecasting used Mean Absolute Percentage Error (MAPE) and to measure the reliability of forecasting results is done by tracking signal. Applications are done on Indonesian red onion production from January 2006 to December 2015. Forecasting of both methods in SSA uses window length L = 39 and grouping r = 8. With α = 0.1, β = 0.001 and γ = 0.5, Holt-Winters additive method gives better result with MAPE 13,469% than SSA method. Keywords:

Download Full-text

PERAMALAN PRODUK DOMESTIK BRUTO (PDB) SEKTOR PERTANIAN, KEHUTANAN, DAN ‎PERIKANAN MENGGUNAKAN SINGULAR SPECTRUM ANALYSIS (SSA)

Jurnal Gaussian ◽

10.14710/j.gauss.v8i1.26623 ◽

2019 ◽

Vol 8 (1) ◽

pp. 68-80

Author(s):

Desy Tresnowati Hardi ◽

Diah Safitri ◽

Agus Rusgiyono

Keyword(s):

Gross Domestic Product ◽

Spectrum Analysis ◽

Time Series Data ◽

Seasonal Pattern ◽

Singular Spectrum Analysis ◽

Domestic Product ◽

Series Data ◽

Percentage Error ◽

Singular Spectrum ◽

Tracking Signal

Forecasting is the process of estimating conditions in the future by testing conditions from the past. One of the forecasting methods is Singular Spectrum Analysis (SSA) which aim of SSA is to make a decomposition of the original series into the sum of a small number of independent and interpretable components such as a slowly varying trend, oscillatory components and a structureless noise. Gross Domestic Product data in the agriculture, forestry, and fisheries sector are time series data with trend and seasonal pattern so that it can be processed using the SSA method. The forecasting process of SSA method uses the main parameter (L) of 21 obtained by the Blind Source Separation (BSS) method. From forecasting, acquired group of 3 groups. Forecasting resulted the value of Mean Absolute Percentage Error (MAPE) is 1.59% and the value of tracking signal is 2.50, which indicates that the results of forecasting is accurate. Keywords: Forecasting, Gross Domestic Product in the agriculture, forestry, and fisheries sector, Singular Spectrum Analysis (SSA)

Download Full-text

PERBANDINGAN KEAKURATAN HASIL PERAMALAN PRODUKSI BAWANG MERAH METODE HOLT-WINTERS DENGAN SINGULAR SPECTRUM ANALYSIS (SSA)

Jurnal Matematika MANTIK ◽

10.15642/mantik.2017.3.1.13-24 ◽

2017 ◽

Vol 3 (1) ◽

pp. 13

Author(s):

Yogo Aryo Jatmiko ◽

Rini Luciani Rahayu ◽

Gumgum Darmawan

Keyword(s):

Spectrum Analysis ◽

Singular Spectrum Analysis ◽

Mean Absolute Percentage Error ◽

Percentage Error ◽

Absolute Percentage Error ◽

Singular Spectrum ◽

Tracking Signal

Metode Holt-Winters digunakan untuk memodelkan data dengan pola musiman, baik mengandung trend maupun tidak. Terdapat dua metode peramalan dalam Singular Spectrum Analysis (SSA), yaitu metode rekuren (R-forecasting) dan metode vektor (V-forecasting). Metode rekuren melakukan kontinuasi secara langsung (dengan bantuan LRF), sedangkan metode vektor berhubungan dengan L-continuation. Perbedaan metode tentunya memberikan perbedaan dalam keakuratan hasil ramalan. Untuk melihat perbedaan antara ketiga metode tersebut dilakukan dengan melihat perbandingan keakuratan dan keandalan hasil ramalan. Untuk mengukur ketepatan peramalan digunakan Mean Absolute Percentage Error (MAPE) dan untuk mengukur keandalan hasil peramalan dilakukan dengan tracking signal. Aplikasi dilakukan pada produksi bawang merah Indonesia periode Januari 2006-Desember 2015. Peramalan kedua metode di SSA menggunakan window length L=39 dan grouping r=8. Dengan nilai α = 0.1, β= 0.001 dan γ=0.5, metode Holt-Winters additive memberikan hasil yang lebih baik dengan MAPE 13,469% dibanding metode SSA.

Download Full-text

PREDIKSI OUTFLOW UANG KARTAL DI KALIMANTAN BARAT DENGAN METODE SINGULAR SPECTRUM ANALYSIS (SSA)

Bimaster : Buletin Ilmiah Matematika, Statistika dan Terapannya ◽

10.26418/bbimst.v8i3.33903 ◽

2019 ◽

Vol 8 (3) ◽

Author(s):

Ageng Wicaksono, Helmi, Yundari

Keyword(s):

Spectrum Analysis ◽

Singular Spectrum Analysis ◽

Mean Absolute Percentage Error ◽

Percentage Error ◽

Absolute Percentage Error ◽

Singular Spectrum ◽

Tracking Signal

Uang merupakan komponen penting dalam perekonomian yang digunakan sebagai alat pembayaran dalam melakukan transaksi jual dan beli. Banyaknya uang yang beredar di masyarakat berpengaruh pada kondisi perekonomian suatu negara. Dalam kewenangan untuk mengatur transaksi arus keluar/masuk uang kartal, diperlukan Rencana Kebutuhan Uang (RKU). Banyak aspek yang menjadi dasar pembuatan RKU, salah satunya data historis aliran uang masuk (inflow) dan aliran uang keluar (outflow). Agar perencanaan tersebut tepat maka peramalan outflow sangat diperlukan. Data outflow pada umumnya memuat pola musiman dan trend sehingga dapat dimodelkan dengan analisis deret waktu. Metode SSA merupakan metode yang menggunakan pendekatan nonparametrik. Artinya dalam pengaplikasiannya, metode ini tidak membutuhkan uji asumsi-asumsi parametrik. Metode SSA menguraikan data deret waktu ke dalam komponen-komponen, yaitu trend, musiman, siklis dan noise. Pada penelitian ini metode SSA yang digunakan adalah metode recurrent (R-forecasting). Penelitian ini dilakukan dengan mengelompokkan data sebanyak 3 kelompok yang menggambarkan fluktuasi data untuk 9 periode peramalan. Tingkat akurasi peramalan diukur menggunakan kriteria Mean Absolute Percentage Error (MAPE) dan Tracking Signal. MAPE dan Tracking Signal pada peramalan SSA terhadap data aktual sebesar 18,63% dan 0,6377. Hal ini mengakibatkan metode SSA dapat dikatakan baik dan valid untuk meramalkan outflow uang kartal di Kalimantan Barat periode November 2018 sampai Juli 2019.Kata Kunci: singular spectrum analysis, R-forecasting, MAPE, tracking signal

Download Full-text

FORECASTING DATA RUNTUN WAKTU MUSIMAN MENGGUNAKAN METODE SINGULAR SPECTRUM ANALYSIS (SSA)

Teorema Teori dan Riset Matematika ◽

10.25157/teorema.v5i1.3286 ◽

2020 ◽

Vol 5 (1) ◽

pp. 47

Author(s):

Dadang Ruhiat ◽

Dini Andiani ◽

Wulan Nurul Kamilah

Keyword(s):

Spectrum Analysis ◽

Moving Average ◽

Singular Spectrum Analysis ◽

Mean Absolute Percentage Error ◽

Percentage Error ◽

Forecasting Model ◽

Autoregressive Integrated Moving Average ◽

Absolute Percentage Error ◽

Singular Spectrum

Pemodelan dan forecasting data runtun waktu akhir-akhir ini terus berkembang dan digunakan di berbagai bidang termasuk di bidang hidrologi. Parameter hidrologi yang sangat penting adalah debit sungai di Indonesia sebagaimana halnya di negara tropis lainnya, besaran dan fluktuasinya dipengaruhi oleh dua faktor musiman, yaitu musim hujan dan kemarau. Pemodelan dan forecasting runtun waktu berbasis statistik pada dasarnya dapat dilakukan melalui dua pendekatan, yaitu statistik parametrik dan statistik non-parametrik. Namun fakta menunjukkan pemodelan dan forecasting runtun waktu melalui pendekatan statistik parametrik lebih banyak dilakukan. Pada penelitian ini dilakukan pemodelan dan forecasting data runtun waktu debit sungai melalui pendekatan statistik non-parametrik dengan menggunakan metode Singular Spectrum Analysis (SSA). Tujuan dari penelitian ini adalah untuk mengetahui hasil forecasting metode SSA dan mengetahui hasil komparasinya dengan hasil forecasting statistik parametrik yang telah dilakukan sebelumnya melalui model Seasonal Autoregressive Integrated Moving Average (SARIMA). Metode proses penelitian ini adalah berupa kajian teori yang kemudian dilanjutkan dengan proses komputasi. Hasil analisis menunjukkan metode SSA memberikan hasil forecasting dengan nilai Mean Absolute Percentage Error (MAPE) yang lebih kecil dari model SARIMA. Dengan demikian disimpulkan forecasting runtun waktu debit sungai yang berpola musiman melalui metode SSA relatif lebih baik dari hasil forecasting model SARIMA.

Download Full-text

Vector Method for Singular Spectrum Analysis (SSA) of Electricity Consumption

2020 International Multi-Conference on Industrial Engineering and Modern Technologies (FarEastCon) ◽

10.1109/fareastcon50210.2020.9271471 ◽

2020 ◽

Author(s):

D.V. Antonenkov ◽

D.A. Pavluchenko ◽

D.V. Orlov ◽

Denis B. Solovev

Keyword(s):

Spectrum Analysis ◽

Singular Spectrum Analysis ◽

Electricity Consumption ◽

Vector Method ◽

Singular Spectrum

Download Full-text

Inflation Forecasting for East Kalimantan Province Using Hybrid Singular Spectrum Analysis- Autoregressive Integrated Moving Average Model

Jurnal Matematika Statistika dan Komputasi ◽

10.20956/j.v18i1.14284 ◽

2021 ◽

Vol 18 (1) ◽

pp. 78-92

Author(s):

Melisa Arumsari ◽

Sri Wahyuningsih ◽

Meiliyani Siringoringo

Keyword(s):

Hybrid Model ◽

Spectrum Analysis ◽

Moving Average ◽

Singular Spectrum Analysis ◽

Percentage Error ◽

Forecasting Accuracy ◽

Autoregressive Integrated Moving Average ◽

Singular Spectrum ◽

East Kalimantan ◽

Inflation Forecasting

The Singular Spectrum Analysis (SSA)-Autoregressive Integrated Moving Average (ARIMA) hybrid method is a good combination of forecasting methods to improve forecasting accuracy and is suitable for economic data that tends to have trend and seasonal patterns, one of which is inflation data. The purpose of this study is to obtain the results of inflation forecasting for East Kalimantan Province in 2021 using the SSA-ARIMA hybrid model. The results of the inflation forecasting for East Kalimantan Province in 2021 using the SSA-ARIMA(1,1,1) hybrid model overall experienced an increase and the highest inflation in 2021 occurred in December of 0.92% with a forecasting accuracy level based on the Root Mean Square Error (RMSE) was 0.069399 and Mean Absolute Percentage Error (MAPE) was 32.61084%

Download Full-text

Sinusoidal Noise Removal in PD Measurement Based on Synchrosqueezing Transform and Singular Spectrum Analysis

Energies ◽

10.3390/en14237967 ◽

2021 ◽

Vol 14 (23) ◽

pp. 7967

Author(s):

Shaorui Qin ◽

Siyuan Zhou ◽

Taiyun Zhu ◽

Shenglong Zhu ◽

Jianlin Li ◽

...

Keyword(s):

White Noise ◽

Spectrum Analysis ◽

Singular Spectrum Analysis ◽

Noise Removal ◽

Time Frequency ◽

Singular Spectrum ◽

Frequency Representation ◽

Synchrosqueezing Transform ◽

Analytic Wavelet ◽

The Difference

In electrical engineering, partial discharge (PD) measurement has been widely used for inspecting and judging insulation conditions of high voltage (HV) apparatus. However, on-site PD measurement easily becomes contaminated by noises. Particularly, sinusoidal noise makes it difficult to recognize real PD signal, thus leading to the misjudgment of insulation conditions. Therefore, sinusoidal noise removal is necessary. In this paper, instantaneous frequency (IF) is introduced, and the synchrosqueezing transform (SST) as well as singular spectrum analysis (SSA) is proposed for sinusoidal noise removal. A continuous analytic wavelet transform is firstly applied to the noisy PD signal and then the time frequency representation (TFR) is reassigned by SST. Narrow-band sinusoidal noise has fixed IF, while PD signal has much larger frequency range and time-varying IF. Due to the difference, the reassigned TFR enables the sinusoidal noise to be distinguished from PD signal. After synthesizing the signal with the recognized IF, SSA is further applied to signal refinement. At last, a numerical simulation is carried out to verify the effectiveness of the proposed method, and its robustness to white noise is also validated. After the implementation of the proposed method, wavelet thresholding can be further applied for white noise reduction.

Download Full-text