Zachiroh, Azmil (2015) Algoritma Forward-Backward dalam Hidden Markov model untuk menganalisis tren sasar saham di Bursa Efek: Studi kasus pada PT Astra Agro Lestari, Tbk. Undergraduate thesis, Universitas Islam Negeri Maulana Malik Ibrahim.
|
Text (Full text)
11610015.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (3MB) | Preview |
Abstract
INDONESIA :
Hidden Markov Model (HMM) dapat diselesaikan dengan berbagai macam algoritma, di antaranya algoritma forward-backward. Parameter HMM ada lima, yaitu jumlah keadaan tersembunyi (N), jumlah keadaan yang teramati (M), matriks peluang transisi (A), matriks peluang observasi (B), dan matriks distribusi awal (π). Apabila nilai dari parameter HMM sudah didapatkan maka langkah selanjutnya adalah pencarian nilai probabilitas dan keadaan yang paling optimal dengan algoritma forward-backward. Algoritma forward dan algoritma backward masing-masing terdiri dari tiga tahap, yaitu inisialisasi, induksi, dan terminasi.
Tujuan penelitian ini adalah memprediksi saham pada periode selanjutnya. Kemudian prediksi saham tersebut dianalisis dan dijadikan suatu keputusan oleh investor. Hasil penelitian ini adalah harga ramalan saham yang didapat selama 15 hari yang cenderung konstan, karena kenaikan dan penurunan harga saham tidak terlalu tinggi. Pergerakan tren pasar optimal yang didapatkan terdiri dari tiga objek, yaitu bullish, bearish, dan sideway. Apabila tren pasar dalam keadaan bullish maka posisi yang baik adalah menjual saham tetapi apabila tren pasar dalam bearish maka posisi yang baik adalah menunggu harga saham kembali naik agar tidak mengalami kerugian.
ENGLISH :
Hidden Markov Model (HMM) can be solved with various algorithms, one of these is forward-backward algorithm. There are five parameters of HMM, namely number of hidden state (N), number of observed state (M), transition probability matrix (A), observation probability matrix (B), and initial distribution matrix (π). If the value of the parameters HMM has been obtained, the next step is the determination of probability value and the state of the most optimal with forward-backward algorithms. Forward algorithm and backward algorithm consist of three stages respectively, namely initialization, induction, and termination.
The aim of this research is predicting stock in the next period. Then the stock prediction were analyzed and a decision by investors was made. The results of this research is the price of forecast stock obtained for 15 days tending constant, because the increase and decrease of stock price not too high. The movement of the market trends is optimal that is obtained consist of three objects, they are bullish, bearish and sideway. If trend market is bullish then a better position is to sell stock but if trend market is bearish then waiting for the stock price to go up is the better choise in order not to suffer losses.
Item Type: | Thesis (Undergraduate) | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Supervisor: | Aziz, Abdul and Barizi, Ahmad | |||||||||
Contributors: |
|
|||||||||
Keywords: | Hidden Markov Model; Algoritma Forward-Backward; Saham; Forward-Backwatd Algorithms; Stock | |||||||||
Subjects: | 01 MATHEMATICAL SCIENCES > 0104 Statistics > 010401 Applied Statistics 01 MATHEMATICAL SCIENCES > 0104 Statistics > 010405 Statistical Theory |
|||||||||
Departement: | Fakultas Sains dan Teknologi > Jurusan Matematika | |||||||||
Depositing User: | M. Muzakir | |||||||||
Date Deposited: | 26 Apr 2017 13:14 | |||||||||
Last Modified: | 26 Apr 2017 13:14 | |||||||||
URI: | http://etheses.uin-malang.ac.id/id/eprint/6369 |
Downloads
Downloads per month over past year
Actions (login required)
View Item |