Analisis Peramalan Penerimaan Negara Bukan Pajak Menggunakan Metode Grey-Markov Dan ANFIS

  • I Dewa Nyoman Anom Manuaba
  • Made Sudarma
  • Nyoman Gunantara

Abstract

Every year, Udayana University publishes the Non-Tax State Revenue Target (TPNBP) as a guide for allocating budgetary funds for the T+2 year. The provision of budget allocations for the T+2 year will depend on how accurately the TPNBP is calculated, which will boost efficiency in carrying out budget planning. Due to the limits of the PNBP realization data at Udayana University and the benefits of the ANFIS approach for forecasting stationary data types, it is required to evaluate the merits of the Gray-Markov and ANFIS methods in order to determine which method is more effective. The study's findings reveal that the Grey-Markov approach yields a value of 0.118% while the ANFIS method yields a value of 4.978%. This suggests that while both methods produce extremely precise results, the Grey-Markov method is more accurate than the ANFIS method due to its smaller MAPE value. The MAPE value for the Grey-Markov approach for the SPI acceptance variable is 0.319%, whereas the MAPE value for the ANFIS method is 23.39%, indicating that the Grey-Markov method is more accurate at predicting SPI acceptance.


Keywords : ANFIS; Grey-Markov; Forecasting; PNBP

Downloads

Download data is not yet available.

References

[1] A. Ahdika, “Model Grey (1,1) dan Grey-Markov pada Peramalan Realisasi Penerimaan Negara,” J. Fourier, vol. 7, no. 1, pp. 1–12, 2018, doi: 10.14421/fourier.2018.71.1-12.
[2] L. Latipah, S. Wahyuningsih, and S. Syaripuddin, “Peramalan Pendapatan Asli Daerah Provinsi Kalimantan Timur Menggunakan Model Grey-Markov (1,1),” Jambura J. Math., vol. 1, no. 2, pp. 89–103, 2019, doi: 10.34312/jjom.v1i2.2347.
[3] Ida Ayu Masyuni, “Peramalan Menggunakan Metode Holt-Winters Untuk Pengujian Kendaraan Bermotor (Studi Kasus: Pengujian Kendaraan Bermotor Kabupaten Tabanan),” J. SPEKTRUM, vol. 6, no. 3, pp. 27–34, 2019.
[4] A. Aryati, I. Purnamasari, and Y. N. Nasution, “Peramalan dengan Menggunakan Metode Holt-Winters Exponential Smoothing (Studi Kasus: Jumlah Wisatawan Mancanegara yang Berkunjung Ke Indonesia) Forecasting using the method of Holt-Winters Exponential Smoothing (Case Study: Number of Foreign Tourists Visi,” J. EKSPONENSIAL, vol. 11, no. 1, pp. 99–105, 2020.
[5] F. A. Widjajati and E. Fani, “Menentukan Penjualan Produk Terbaik di Perusahaan X Dengan Metode Winter Eksponensial Smoothing Dan Metode Event Based,” Limits J. Math. Its Appl., vol. 14, no. 1, pp. 25–35, 2017, [Online]. Available: http://iptek.its.ac.id/index.php/limits/article/view/2127.
[6] W. Handoko, “Prediksi Jumlah Penerimaan Mahasiswa Baru Dengan Metode Single Exponential Smoothing (Studi Kasus: Amik Royal Kisaran),” JURTEKSI (Jurnal Teknol. dan Sist. Informasi), vol. 5, no. 2, pp. 125–132, 2019, doi: 10.33330/jurteksi.v5i2.356.
[7] Y. Asri and D. Permana, “Peramalan Penerimaan Pajak Negara Indonesia Tahun 2019 Menggunakan Metode Pemulusan Eksponensial Ganda Tipe Brown,” UNP J. Math., pp. 70–74, 2019, [Online]. Available: http://ejournal.unp.ac.id/students/index.php/mat/article/view/6321%0Ahttp://ejournal.unp.ac.id/students/index.php/mat/article/viewFile/6321/3228.
[8] I. G. N. R. D. Widhura, M. Sudarma, and R. S. Hartati, “Penentuan Target Pajak Kendaraan Bermotor Di Provinsi Bali Menggunakan ARIMA Dan Algoritma Genetik,” Maj. Ilm. Teknol. Elektro, vol. 17, no. 3, pp. 345–352, 2018, doi: 10.24843/mite.2018.v17i03.p07.
[9] Y. Pramana, R. S. Hartati, and K. Oka Saputra, “Peramalan Penerbitan Ijin Mendirikan Bangunan Dengan Single Moving Average Dan Exponential Smoothing,” Maj. Ilm. Teknol. Elektro, vol. 18, no. 2, pp. 241–248, 2019, doi: 10.24843/mite.2019.v18i02.p13.
[10] T. N. Putri, A. Yordan, and D. H. Lamkaruna, “Peramalan Penerimaan Mahasiswa Baru Universitas Samudra Menggunakan Metode Regresi Linear Sederhana,” J-TIFA (Jurnal Teknol. Inform., vol. 2, no. 1, pp. 21–27, 2019.
[11] S. Sutrisman, H. Syafwan, and ..., “Implementation of Trend Moment Method in Forecasting Regional Income,” Build. Informatics, Technol. Sci., vol. 4, no. 2, pp. 749–758, 2022, doi: 10.47065/bits.v4i2.2090.
[12] F. P. Wahyu, I. D. Rahmawati, and K. Umam, “Identifying Best Method for Forecasting Tax Income using Time Series Analysis,” IAPA Int. Conf. Int. Indones. Conf. Interdiscip. Stud., vol. Dec, pp. 60–73, 2022.
[13] I. Fitria, M. S. K. Alam, and S. Subchan, “Perbandingan Metode ARIMA dan Exponential Smoothing pada Peramalan Harga Saham LQ45 Tiga Perusahaan dengan Nilai Earning Per Share (EPS) Tertinggi,” Limits J. Math. Its Appl., vol. 14, no. 2, pp. 113–125, 2017, doi: 10.12962/limits.v14i2.3060.
[14] Y. Surya Bhakti, A. Budiman Kusdinar, D. Asril, and A. Sunarto, “Model Peramalan Penerimaan Calon Mahasiswa Menggunakan Metode Regresi,” J. Ilm. Komput., vol. 16, no. 2, pp. 113–120, 2020.
[15] A. R. Lasri Nijal, Roki Hardianto, “Peramalan Penerimaan Karyawan PT . Cipta Persada Infrastruktur Menggunakan Monte Carlo,” J. Sist. Inf., vol. 3, no. 2, pp. 98–115, 2020.
[16] D. A. Setiawan, S. Wahyuningsih, and R. Goejantoro, “Peramalan Produksi Kelapa Sawit Menggunakan Winter’s dan Pegel’s Exponential Smoothing dengan Pemantauan Tracking Signal,” Jambura J. Math., vol. 2, no. 1, pp. 1–14, 2020, doi: 10.34312/jjom.v2i1.2320.
[17] I. D. N. Anom Manuaba, I. B. Gede Manuaba, and M. Sudarma, “Komparasi Metode Peramalan Grey dan Grey- Markov untuk mengetahui Peramalan PNBP di Universitas Udayana,” vol. 21, no. 1, pp. 83–88, 2022.
[18] A. Muqtadir, S. Suryono, and V. Gunawan, “The Implementation of Grey Forecasting Model for Forecast Result’s Food Crop Agricultural,” Sci. J. Informatics, vol. 3, no. 2, pp. 159–166, 2016, doi: 10.15294/sji.v3i2.7912.
[19] I. Sidiq, E. Febianti, and P. F. Ferdinant, “Peramalan Kebutuhan Konsumsi Listrik Menggunakan Grey Prediction Model,” J. Tek. Ind., vol. 1, no. 1, pp. 1–6, 2013, [Online]. Available: http://jurnal.untirta.ac.id/index.php/jti/article/view/109.
[20] I. N. Z. S. M. Nurfitri Imro’ah, “Peramalan Harga Emas Batangan Menggunakan Metode Grey Double Exponential Smoothing,” Bimaster Bul. Ilm. Mat. Stat. dan Ter., vol. 9, no. 4, pp. 497–504, 2020, doi: 10.26418/bbimst.v9i4.42280.
[21] G. F. Fitri, F. Agustina, and R. Marwati, “Penerapan Metode Grey System Pada Peramalan Produk Olefins (Studi Kasus PT. Chandra Asri Petrochemical Tbk),” EurekaMatika, vol. 6, no. 2, pp. 52–63, 2018.
[22] A. Fitro, Rudianto, and H. Prasetyo, “Implementasi Metode Grey Verhulst Untuk Mendukung Kebijakan Dalam Mengantisipasi Mahasiswa Dropout,” vol. 3, no. 02, pp. 180–187, 2021.
[23] L. D. Immawan and A. Ahdika, “Comparison of Grey-Markov (1,1), Grey-Markov (2,1), and moving average methods in forecasting small sized data of the unit price of materials in batam,” AIP Conf. Proc., vol. 2021, no. 2018, 2018, doi: 10.1063/1.5062783.
[24] N. L. Nariswari and C. N. Rosyidi, “Aplikasi Metode Grey Forecasting Pada Peramalan Kebutuhan Bahan Bakar Alternatif Ramah Lingkungan di PT. Indocement Tunggal Prakarsa Tbk,” PERFORMA, vol. 14, no. 2, pp. 99–106, 2015, doi: 10.20961/performa.14.2.10986.
[25] D. R. Darmawanti, Yundari, and N. M. Huda, “Prediksi Realisasi Penerimaan Pajak Bumi Dan Bangunan Provinsi Kalimantan Barat Dengan Model Grey-Markov(1, 1),” Bul. Ilm. Math. Stat dan Ter., vol. 11, no. 3, pp. 503–512, 2022.
[26] B. H. S. Atma and S. Sugiyarto, “Adaptive neuro fuzzy inference system untuk peramalan jumlah wisatawan,” J. Ilm. Mat., vol. 7, no. 1, p. 1, 2020, doi: 10.26555/konvergensi.v7i1.19195.
[27] I. G. B. Ngurah Diksa, “Peramalan Gelombang Covid 19 Menggunakan Hybrid Nonlinear Regression Logistic – Double Exponential Smoothing di Indonesia dan Prancis,” Jambura J. Math., vol. 3, no. 1, pp. 37–51, 2021, doi: 10.34312/jjom.v3i1.7771.
[28] H. W. Tresnani, A. Sihabuddin, and K. Mustofa, “Optimasi Parameter Pada Metode Peramalan Grey Holt - Winter Exponential Smoothing Dengan Golden Section,” Berk. MIPA, vol. 25, no. 3, pp. 312–325, 2018.
[29] D. I. Purnama and O. P. Hendarsin, “Peramalan Jumlah Penumpang Berangkat Melalui Transportasi Udara di Sulawesi Tengah Menggunakan Support Vector Regression (SVR),” Jambura J. Math., vol. 2, no. 2, pp. 49–59, 2020, doi: 10.34312/jjom.v2i2.4458.
[30] D. N. Adli, “Prediksi Harga Jagung Menggunakan Metode Fuzzy Time Series Dengan Atau Tanpa Menggunakan Markov Chain,” J. Nutr. Ternak Trop., vol. 4, no. 1, pp. 49–54, 2021, doi: 10.21776/ub.jnt.2021.004.01.6.
[31] R. Assakhiy, S. Anwar, and F. Ar, “Peramalan Realisasi Penerimaan Zakat Pada Baitulmal Aceh Dengan Mempertimbangkan Efek Dari Variasi Kalender,” J. Ekon. dan Pembang., vol. 27, no. 2, pp. 27–45, 2019.
[32] L. H. Zulkieflimansyah, Muhammad Nurjihadi, Diah Anggeraini Hasri, Nova Adhitya Ananda, “Proyeksi Dinamika Pendapatan Asli Daerah Nusa Tenggara Barat,” J. Maneksi, vol. 10, no. 2, pp. 231–245, 2021.
[33] G. B. Nabilah, Y. N. Nasution, and I. Purnamasari, “Peramalan Indeks Harga Konsumen Provinsi Kalimantan Timur dengan Metode Grey Double Exponential Smoothing Holt,” Pros. Semin. Nas. Mat. Stat. dan Apl., vol. Mei, no. II, pp. 69–80, 2022.
[34] I. G. D. A. L K Widyapratiwi, I P A Mertasana, “Peramalan Beban Listrik Jangka Pendek Di Bali Menggunakan Pendekatan Adaptive Neuro-Fuzzy Inference System (Anfis),” Maj. Ilm. Teknol. Elektro, vol. 11, no. 2, 2013.
[35] C. B. Sinaga, H. Haviluddin, H. S. Pakpahan, A. Prafanto, and H. J. Setyadi, “Peramalan Curah Hujan Dengan Pendekatan Adaptive Neuro Fuzzy Inference System,” Sains, Apl. Komputasi dan Teknol. Inf., vol. 1, no. 2, p. 1, 2019, doi: 10.30872/jsakti.v1i2.2599.
Published
2023-06-05
How to Cite
MANUABA, I Dewa Nyoman Anom; SUDARMA, Made; GUNANTARA, Nyoman. Analisis Peramalan Penerimaan Negara Bukan Pajak Menggunakan Metode Grey-Markov Dan ANFIS. Majalah Ilmiah Teknologi Elektro, [S.l.], v. 22, n. 1, p. 1 - 10, june 2023. ISSN 2503-2372. Available at: <https://ojs.unud.ac.id/index.php/jte/article/view/94501>. Date accessed: 25 apr. 2024. doi: https://doi.org/10.24843/MITE.2023.v22i01.P01.