ANALISIS DOUBLE EXPONENTIAL SMOOTHING HOLT BERBASIS ALGORITMA KUADRATIK DAN MODEL SARIMAX DENGAN VARIASI KALENDER DALAM PERAMALAN INFLASI INDONESIA

ANDHIKA RANGGA PUTRA ISCHAQ, . (2026) ANALISIS DOUBLE EXPONENTIAL SMOOTHING HOLT BERBASIS ALGORITMA KUADRATIK DAN MODEL SARIMAX DENGAN VARIASI KALENDER DALAM PERAMALAN INFLASI INDONESIA. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text Cover__.pdf Download (1MB)
	Text BAB 1.pdf Download (296kB)
	Text BAB 2.pdf Restricted to Registered users only Download (738kB) | Request a copy
	Text BAB 3.pdf Restricted to Registered users only Download (365kB) | Request a copy
	Text BAB 4.pdf Restricted to Registered users only Download (955kB) | Request a copy
	Text BAB 5.pdf Restricted to Registered users only Download (312kB) | Request a copy
	Text Daftar Pustaka.pdf Download (242kB)
	Text Lampiran.pdf Restricted to Registered users only Download (669kB) | Request a copy

Abstract

Inflasi merupakan salah satu indikator ekonomi makro yang sangat dinamis dan sensitif terhadap guncangan eksternal, termasuk pergeseran konsumsi masyarakat akibat pengaruh musiman keagamaan dan akhir tahun. Penelitian ini bertujuan untuk menganalisis penerapan dua pendekatan matematis dalam peramalan jangka pendek inflasi Indonesia, yaitu metode Double Exponential Smoothing (DES) Holt berbasis Algoritma Kuadratik dan model Seasonal Autoregressive Integrated Moving Average with Exogenous Variables (SARIMAX) dengan variasi kalender. Pendekatan pertama berfokus pada optimasi parameter internal model melalui penerapan metode Double Exponential Smoothing Holt menggunakan Algoritma Kuadratik untuk memperoleh parameter pemulusan yang optimal. Pendekatan kedua memanfaatkan model Seasonal Autoregressive Integrated Moving Average with Exogenous Variables (SARIMAX) dengan mengintegrasikan variabel variasi kalender, yaitu satu bulan sebelum Idulfitri, bulan Idulfitri, serta periode Natal dan Tahun Baru dalam bentuk variabel dummy. Data yang digunakan merupakan data tingkat inflasi bulanan Indonesia periode Mei 2016 sampai Maret 2026, yang dibagi menjadi 113 data pelatihan (training) dan 6 data pengujian (testing). Akurasi peramalan dievaluasi menggunakan Mean Absolute Percentage Error (MAPE). Hasil analisis menunjukkan bahwa parameter optimal metode DES Holt berbasis Algoritma Kuadratik adalah alpha = 0,9999 dan gamma = 0,7133, dengan nilai MAPE sebesar 13,5931% pada data pelatihan dan 13,2044% pada data pengujian. Sementara itu, model SARIMAX yang terpilih adalah SARIMAX(0,1,1)(0,0,1){12} berdasarkan evaluasi nilai Akaike Information Criterion (AIC), signifikansi parameter, diagnostik residual, dan akurasi peramalan. Model tersebut menghasilkan nilai MAPE sebesar 14,0863% pada data pelatihan dan 3,8435% pada data pengujian. Berdasarkan hasil analisis pada data penelitian ini, model SARIMAX menghasilkan tingkat akurasi peramalan yang lebih tinggi dibandingkan metode DES Holt berbasis Algoritma Kuadratik. Hasil tersebut menunjukkan bahwa pemanfaatan informasi eksternal berupa variasi kalender berpotensi meningkatkan akurasi peramalan inflasi pada data yang dianalisis.***** Inflation is one of the most dynamic macroeconomic indicators and is highly sensitive to external shocks, including changes in public consumption patterns caused by religious seasonal events and the year-end period. This study aims to analyze the application of two mathematical approaches for short-term forecasting of Indonesia's inflation, namely the Quadratic Algorithm-based Double Exponential Smoothing (DES) Holt method and the Seasonal Autoregressive Integrated Moving Average with Exogenous Variables (SARIMAX) model incorporating calendar variations. The first approach focuses on optimizing the internal parameters of the Double Exponential Smoothing Holt method using the Quadratic Algorithm to obtain optimal smoothing parameters. The second approach employs the Seasonal Autoregressive Integrated Moving Average with Exogenous Variables (SARIMAX) model by incorporating calendar variation variables, namely one month before Eid al-Fitr, the Eid al-Fitr month, and the Christmas and New Year period, in the form of dummy variables. The data used in this study consist of monthly inflation data in Indonesia from May 2016 to March 2026, which were divided into 113 training observations and 6 testing observations. Forecasting accuracy was evaluated using the Mean Absolute Percentage Error (MAPE). The results show that the optimal parameters of the Quadratic Algorithm-based DES Holt method are alpha = 0.9999 and gamma = 0.7133, yielding MAPE values of 13.5931% for the training data and 13.2044% for the testing data. Meanwhile, the selected SARIMAX model was SARIMAX(0,1,1)(0,0,1)_12, selected based on the evaluation of the Akaike Information Criterion (AIC), parameter significance, residual diagnostics, and forecasting accuracy. The model produced MAPE values of 14.0863% for the training data and 3.8435% for the testing data. Based on the analysis of the data used in this study, the SARIMAX model achieved higher forecasting accuracy than the Quadratic Algorithm-based DES Holt method. These findings indicate that incorporating external information through calendar variations has the potential to improve the accuracy of inflation forecasting for the data analyzed.

Item Type:	Thesis (Sarjana)
Additional Information:	1. Ibnu Hadi, M.Si. 2. Dr. Eti Dwi Wiraningsih, M.Si.
Subjects:	Sains > Matematika Sains > Statistika
Divisions:	FMIPA > S1 Matematika
Depositing User:	Andhika Rangga Putra Ischaq .
Date Deposited:	18 Aug 2026 04:00
Last Modified:	18 Aug 2026 04:00
URI:	http://repository.unj.ac.id/id/eprint/71482

Actions (login required)

View Item