Please use this identifier to cite or link to this item: http://repository.ipb.ac.id/handle/123456789/115337
Title: Implementasi Algoritma Viterbi untuk Memprediksi Kondisi Pasar Saham
Other Titles: The Implementation of Viterbi Algorithm to Predict Stock Market Conditions
Authors: Setiawaty, Berlian
Lesmana, Donny Citra
Lestari, Amanah Yuni
Issue Date: 2022
Publisher: IPB University
Abstract: Kondisi pasar saham selalu mengalami perubahan setiap waktu. Hal ini ditunjukkan salah satunya oleh pergerakan nilai penutupan suatu indeks saham baik meningkat maupun menurun. Kondisi pasar saham memengaruhi sentimen pasar dalam memutuskan suatu strategi perdagangan, oleh karena itu penting untuk diketahui. Jika kondisi pasar saham yang menyebabkan terjadinya nilai penutupan suatu indeks saham diasumsikan tidak teramati secara langsung dan membentuk rantai Markov, maka pasangan kondisi pasar saham dan nilai penutupan indeks saham dapat dimodelkan sebagai hidden Markov model. Terdapat tiga permasalahan dasar yang harus diselesaikan dalam hidden Markov model yaitu evaluation problem, decoding problem, dan learning problem. Beberapa algoritma yang dapat digunakan untuk menyelesaikan permasalahan tersebut adalah algoritma ¬forward-backward, algoritma Viterbi, dan algoritma Baum-Welch. Saat decoding problem menggunakan algoritma Viterbi inilah kondisi pasar saham yang terjadi dapat diduga. Pada tugas akhir ini akan diduga kondisi pasar saham dari nilai penutupan harian indeks FTSE 100. Dari hasil perhitungan, selama periode pengamatan, nilai penutupan harian indeks FTSE 100 disebabkan oleh tiga state hidden Markov yang dapat dikategorikan sebagai bullish, sideways, dan bearish.
Stock market conditions change over time. It is indicated by the movement of the closing values of a stock index either increasing or decreasing. It affects market sentiment in deciding a trading strategy, therefore it is important to know. If the stock market conditions that cause the closing values of a stock index are assumed not to be observed directly and form a Markov chain, then the pair of stock market conditions and the closing values of a stock index can be modelled as a hidden Markov model. Three basic problems must be solved in hidden Markov model are evaluation problem, decoding problem, and learning problem. Several algorithms that can be used to solve these problems are forward-backward algorithm, Viterbi algorithm, and Baum-Welch algorithm. When the decoding problem using Viterbi algorithm, the stock market conditions that occur can be predicted. In this final project, the stock market conditions will be estimated from the daily closing values of FTSE 100 index. From the calculation results, during the observation period, the daily closing values of FTSE 100 index are caused by three hidden Markov states which can be categorized as bullish, sideways, and bearish.
URI: http://repository.ipb.ac.id/handle/123456789/115337
Appears in Collections:UT - Actuaria

Files in This Item:
File Description SizeFormat 
Cover, Lembar Pengesahan, Prakata, Daftar Isi.pdf
  Restricted Access
Cover603.88 kBAdobe PDFView/Open
G94180004_Amanah Yuni Lestari.pdf
  Restricted Access
Fullteks1.52 MBAdobe PDFView/Open
Lampiran.pdf
  Restricted Access
Lampiran611.72 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.