Prediksi Saham Menggunakan Model Ensemble Berbasis Pohon Keputusan Dalam Kerangka Stacking Yang Mempertimbangkan Indikator Teknikal Dan Sentimen Berita

Wiranata, Rico Bayu (2021) Prediksi Saham Menggunakan Model Ensemble Berbasis Pohon Keputusan Dalam Kerangka Stacking Yang Mempertimbangkan Indikator Teknikal Dan Sentimen Berita. Masters thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
05211950010017-Master_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 October 2023.

Download (4MB) | Request a copy

Abstract

Investor harus memprediksi saham dengan tepat agar keuntungan maksimal sekaligus terhindar kebangkrutan. Namun bursa saham sulit dideteksi karena perilakunya tidak hanya dipengaruhi oleh faktor historis harga dan indikator teknikal, tetapi juga dipengaruhi oleh situasi politik, kinerja perusahaan, maupun perekonomian global yang tersedia melalui berita. Penelitian ini bertujuan mengembangkan model prediksi saham yang mengkombinasikan indikator teknikal saham dan sentimen berita menggunakan metode stacking ensemble berbasis pohon keputusan. Dalam metode ini, beberapa pengklasifikasi berbasis algoritma pohon keputusan digunakan sebagai base-learner yang kemudian ditumpuk menggunakan algoritma stacking sebagai meta-learner. Masing-masing base-learner dibangun melalui proses optimasi hyper-parameter pohon keputusan menggunakan algoritma genetika. Algoritma stacking kemudian digunakan untuk menumpuk hasil prediksi semua base-learner untuk membangun model prediksi akhir. Model prediksi yang dikembangkan dalam penelitian ini diuji coba menggunakan data sahan dari tiga bank nasional utama, yaitu Bank Rakyat Indonesia (BBRI), Bank Mandiri (BMRI), dan Bank Negara Indonesia (BBNI). Uji coba menggunakan data saham ketiga bank tersebut memberikan hasil akurasi berturut-turut sebesar 83,67%; 87,34%; 86,53% dan hasil f1-score berturut-turut sebesar 83,05%; 86,22%; 83,90%. Dampak dari hasil prediksi terhadap evaluasi perdagangan saham BBRI, BMRI, dan BBNI berturut-turut memberikan return sebesar 94,91%; 103,64%, 105,40% dan keuntungan bersih dalam setahun berturut-turut sebesar 102,30%; 122,52%; 119,94%. Hasil evaluasi perdagangan tersebut memberikan nilai maximum drawdown yang cenderung paling kecil dan nilai rasio sharpe yang tinggi. ========================================================================================== Investors must predict stocks correctly in order to maximize their profits while avoiding bankruptcy. However, the stock market is difficult to detect because its behavior is not only influenced by historical price factors and technical indicators, but is also influenced by the political situation, company performance, and the global economy that is available through the news. This study aims to develop a stock prediction model that combines stock technical indicators and related news sentiment using a decision tree-based stacking ensemble method. In this method, several classifiers based on decision tree algorithms are used as base-learners which are then stacked using a stacking algorithm as a meta-learner. Each base-learner is built through a decision tree hyper-parameter optimization process using a genetic algorithm. The stacking algorithm is then used to stack the predicted results of all base-learners to build the final prediction model. The prediction model developed in this study was tested using valid data from three main national banks, namely Bank Rakyat Indonesia (BBRI), Bank Mandiri (BMRI), and Bank Negara Indonesia (BBNI). The experimental results using the stock data of these three banks gave accuracy of 83.67%, 87.34%, 86.53%, respectively; and f1-score results of 83.05%, 86.22%, 83.90%, respectively. The impact of the prediction results on the evaluation of trading shares of BBRI, BMRI, and BBNI respectively gave a return of 94.91%, 103.64%, 105.40% and a year's net profit in a row of 102.30%, 122.52%, 119.94%. Moreover, the results of the trade evaluation provide a maximum drawdown value that tends to be the smallest with a high sharpe ratio value.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Kata kunci: metode ensemble, metode stacking, algoritma genetika, penyetelan hyper-parameter, prediksi saham, indikator teknikal, analisis sentimen Keywords: ensemble method, stacking method, genetic algorithm, stock prediction, hyper-parameter tuning, technical indicators, sentiment analysis
Subjects: H Social Sciences > HB Economic Theory > Economic forecasting--Mathematical models.
H Social Sciences > HD Industries. Land use. Labor > HD108 Classification (Theory. Method. Relation to other subjects )
H Social Sciences > HD Industries. Land use. Labor > HD30.23 Decision making. Business requirements analysis.
H Social Sciences > HD Industries. Land use. Labor > HD38.7 Business intelligence. Trade secrets
H Social Sciences > HG Finance > HG4529 Investment analysis
Q Science > Q Science (General) > Q180.55.M38 Mathematical models
Q Science > Q Science (General) > Q325.5 Machine learning.
Q Science > Q Science (General) > Q370 Entropy (Information theory)
Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Q Science > QA Mathematics > QA278.5 Principal components analysis. Factor analysis. Correspondence analysis (Statistics)
Q Science > QA Mathematics > QA336 Artificial Intelligence
Q Science > QA Mathematics > QA402 System analysis.
Q Science > QA Mathematics > QA402.5 Genetic algorithms.
Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science)
Q Science > QA Mathematics > QA76.9D338 Data integration
Q Science > QA Mathematics > QA9.58 Algorithms
T Technology > T Technology (General) > T174 Technological forecasting
T Technology > T Technology (General) > T57.5 Data Processing
T Technology > T Technology (General) > T57.84 Heuristic algorithms.
T Technology > T Technology (General) > T58.62 Decision support systems
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 59101-(S2) Master Thesis
Depositing User: RICO BAYU WIRANATA
Date Deposited: 19 Aug 2021 14:18
Last Modified: 19 Aug 2021 14:18
URI: https://repository.its.ac.id/id/eprint/88342

Actions (login required)

View Item View Item