View Item 
      •   IPB Repository
      • Dissertations and Theses
      • Undergraduate Theses
      • UT - Faculty of Mathematics and Natural Sciences
      • UT - Actuaria
      • View Item
      •   IPB Repository
      • Dissertations and Theses
      • Undergraduate Theses
      • UT - Faculty of Mathematics and Natural Sciences
      • UT - Actuaria
      • View Item
      JavaScript is disabled for your browser. Some features of this site may not work without it.

      Peramalan Harga Minyak Bumi menggunakan Model Hidden Markov dengan Teknik Smoothing Exponential Moving Average

      Thumbnail
      View/Open
      Cover (773.6Kb)
      Fulltext (2.097Mb)
      Lampiran (984.3Kb)
      Date
      2024
      Author
      Wijaya, Dian Arya Dwi Kurnia
      Setiawaty, Berlian
      Metadata
      Show full item record
      Abstract
      Minyak bumi merupakan hasil sumber daya alam yang memiliki beragam produk turunan dan menjadikannya suatu komoditas alam yang sangat berharga. Minyak bumi juga digunakan sebagai aset finansial yang diperdagangkan di pasar komoditas. Harga minyak bumi cukup fluktuatif sehingga kemampuan dalam peramalam sangat dibutuhkan. Karya ilmiah ini mengasumsikan harga minyak bumi memiliki faktor yang tidak teramati dan membentuk rantai Markov, sehingga model hidden Markov dapat digunakan. Data yang digunakan merupakan data harga penutupan mingguan minyak bumi yang diamati dari 5 Januari 2009 sampai 12 Februari 2024, diperoleh proporsi data training terbaik yaitu 93.5% data atau sebanyak 737 data. Model hidden Markov yang dibentuk menunjukkan adanya tiga state tersembunyi yang secara berurutan memiliki sebaran Weibull(2.56, 22.74, 75.86), Weibull(3.02, 19.19, 55.20), dan Weibull(23.05, 171.74, -119.14). Hasil simulasi diperoleh MAPE data training sebesar 17.78% dan MAPE data testing sebesar 10.02%. Teknik pemulusan exponential moving average dengan ?? = 0.095pada data simulasi dapat menurunkan nilai MAPE menjadi 13.42% pada MAPE training dan 7.05% pada MAPE testing.
       
      Crude oil is a natural resource that yields a variety of derivative products, thus making it a highly valuable natural commodity. Crude oil is also utilized as a financial asset traded in commodity markets. Given the considerable volatility of crude oil prices, the ability to forecast accurately is highly essential. This scientific work assumes that the price of crude oil has unobserved factors and forms a Markov chain. Thus, the Hidden Markov Model can be used. The data used is weekly closing price of crude oil, observed from January 5, 2009, to February 12, 2024, the best proportion of training data obtained is 93.5%, totaling 737 data points. the hidden Markov model formed indicates the presence of three hidden states, sequentially exhibiting Weibull(2.56, 22.74, 75.86), Weibull(3.02, 19.19, 55.20), and Weibull(23.05, 171.74, -119.14). Simulation results yield a MAPE of 17.78% for the training data and a MAPE of 10.02% for the testing data. The exponential moving average smoothing technique with ?? = 0.095 applied to the simulation data can reduce the MAPE to 13.42% for the training set and 7.05% for the testing set.
       
      URI
      http://repository.ipb.ac.id/handle/123456789/155331
      Collections
      • UT - Actuaria [205]

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository
        

       

      Browse

      All of IPB RepositoryCollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

      My Account

      Login

      Application

      google store

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository