Image of Prediksi Harga Saham Menggunakan Pendekatan Machine Learning Dan Deep Learning: Implementasi Knn Dan Lstm Pada Saham Sektor Infrastruktur Dan Energi
Penanda Bagikan

CD Skripsi

Prediksi Harga Saham Menggunakan Pendekatan Machine Learning Dan Deep Learning: Implementasi Knn Dan Lstm Pada Saham Sektor Infrastruktur Dan Energi

High stock price volatility in Indonesia's infrastructure and energy sectors poses challenges in predicting price movements. This study aims to implement and compare the accuracy of K-Nearest Neighbors (KNN) machine learning model and Long Short-Term Memory (LSTM) deep learning model in predicting stock prices in these sectors. The research stages include data preprocessing, model training, evaluation using Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE) metrics, and result analysis. The data used consists of daily closing prices from four selected stocks: PTPP (PT PP Persero Tbk), BUKK (Bukaka Teknik Utama Tbk), BUMI (Bumi Resources Tbk), and MEDC (Medco Energi Internasional Tbk) during the period from January 2022 to December 2024. The total data analyzed amounts to 2,888 rows, with 722 data points per stock. The research results show that LSTM consistently provides more accurate predictions with lower MAPE values compared to KNN. In the energy sector, LSTM recorded MAPE of 4.06% for BUMI and 2.16% for MEDC, while KNN recorded 10.34% and 5.70%, showing a reduction of up to half by LSTM. On the other hand, KNN demonstrated competitive performance on PTPP with RMSE of 31.22 compared to LSTM's 31.76. Therefore, LSTM is more effective for predicting stock prices with high volatility and complex movement patterns, while KNN can serve as a fast and simple alternative prediction model.

Keywords: Stock Price Prediction, KNN, LSTM, Infrastructure Sector, Energy Sector

Ketersediaan
#
Perpustakaan Universitas Riau 2103112028
2103112028
Tersedia
Informasi Detail
Judul Seri
-
No. Panggil
2103112028
Penerbit
Pekanbaru : Universitas Riau FMIPA Sistem Informasi.,
Deskripsi Fisik
-
Bahasa
Indonesia
ISBN/ISSN
-
Klasifikasi
2103112028
Tipe Isi
-
Tipe Media
-
Tipe Pembawa
-
Edisi
-
Subjek
SISTEM INFORMASI
Info Detail Spesifik
-
Pernyataan Tanggungjawab
Versi lain/terkait

Tidak tersedia versi lain

Lampiran Berkas
Komentar
Anda harus masuk sebelum memberikan komentar