Prediksi Harga Saham Menggunakan Analisis Sentimen Berita dan Algoritma Random Forest

Nathaniel, Nathaniel (2024) Prediksi Harga Saham Menggunakan Analisis Sentimen Berita dan Algoritma Random Forest. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5031201062-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (2MB) | Request a copy

Abstract

Harga saham dipengaruhi oleh interaksi kompleks elemen-elemen yang mencakup nilai intrinsik perusahaan dan keadaan ekonomi dan pasar yang lebih luas. Selain itu, sentimen berita juga berkorelasi dengan pergerakan harga saham. Dengan demikian, memprediksi pasar saham adalah tugas yang menantang. Beberapa teknik digunakan untuk mencapai tujuan ini, salah satunya adalah machine learning. Machine learning telah menjadi instrumen yang efektif untuk memprediksi harga saham berkat kemampuannya untuk menganalisis dan mempelajari kumpulan data yang kompleks. Penelitian ini mengintegrasikan analisis sentimen dan machine learning untuk memprediksi harga saham, dengan fokus pada Indeks LQ45 dan berita keuangan dari portal CNN Indonesia, CNBC Indonesia, dan Detik.com. Penelitian ini menggunakan algoritma Random Forest, dengan memasukkan skor sentimen dari artikel berita sebagai fitur. Hasilnya menunjukkan bahwa Random Forest secara konsisten mengungguli model-model regresi lainnya, meskipun akurasinya bervariasi antar saham. Fitur berbasis harga, khususnya 'Low' dan 'High', paling berpengaruh dalam prediksi. Skor sentimen dan volume memiliki dampak minimal ketika digabungkan dengan data harga. Namun, skor sentimen secara signifikan memengaruhi prediksi ketika diisolasi dari fitur-fitur yang berhubungan dengan harga.

=======================================================================================================================

Stock prices are influenced by a complex interaction of elements encompassing the company's intrinsic value and the broader economic and market circumstances. Other than that, news sentiment also correlates with stock price movements. Thus, Predicting the stock market is undeniably a compelling task. Several techniques are employed to achieve this objective, one of which is machine learning. Machine learning has become an effective instrument for predicting stock prices thanks to its ability to analyze and learn from complex datasets. This research integrates sentiment analysis and machine learning to predict stock prices, focusing on the LQ45 Index and financial news from Indonesian portals CNN Indonesia, CNBC Indonesia, and Detik.com. The study employs the Random Forest algorithm, incorporating sentiment scores from news articles as a feature. Results show that Random Forest consistently outperformed other regression models, though accuracy varied across stocks. Price-based features, particularly 'Low' and 'High', were most influential in predictions. Sentiment scores and volume had minimal impact when combined with price data. However, sentiment scores significantly influenced predictions when isolated from price-related features.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Algoritma Random Forest, Analisis Sentimen, Berita Financial, Machine Learning, Prediksi Harga Saham ===================================================================================================================================== Financial News, Machine Learning, Random Forest Algorithm, Sentiment Analysis, Stock Price Prediction
Subjects:	H Social Sciences > HG Finance > HG4915 Stocks--Prices Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Divisions:	Faculty of Business and Management Technology > Business Management > 61205-(S1) Undergraduate Thesis
Depositing User:	Nathaniel Nathaniel
Date Deposited:	05 Aug 2024 08:54
Last Modified:	05 Aug 2024 08:54
URI:	http://repository.its.ac.id/id/eprint/111426

Actions (login required)

View Item