Analisis Prediktif Harga Saham Menggunakan Model Convolutional Neural Network Long Short Term Memory = Predictive Analysis Of Stock Price Using Convolutional Neural Network Long Short Term Memory Model

Widjaya, Jeffrey (2023) Analisis Prediktif Harga Saham Menggunakan Model Convolutional Neural Network Long Short Term Memory = Predictive Analysis Of Stock Price Using Convolutional Neural Network Long Short Term Memory Model. Bachelor thesis, Universitas Pelita Harapan.

[img] Text (Title)
Title.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (142kB)
[img] Text (Abstract)
Abstract.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (588kB)
[img] Text (ToC)
ToC.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB)
[img] Text (Chapter1)
Chapter 1.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (576kB)
[img] Text (Chapter2)
Chapter 2.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB)
[img] Text (Chapter3)
Chapter 3.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (686kB)
[img] Text (Chapter4)
Chapter 4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (687kB)
[img] Text (Chapter5)
Chapter 5.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (570kB)
[img] Text (Bibliography)
Bibliography.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB)
[img] Text (Appendices)
Appendices.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (23MB)

Abstract

Kemajuan teknologi memungkinkan untuk memprediksi harga saham dengan menggunakan metode machine learning. Metode deep learning yang akan digunakan adalah Convolutional Neural Network Long Short Term Memory dan Convolutional Neural Network Bidirectional Long Short Term Memory. Tujuan dari penelitian ini adalah mengetahui dampak dari jumlah layer terhadap model dan menggunakan metriks evaluasi untuk menentukan model yang lebih baik dalam memprediksi harga saham. Model tersebut adalah model CNN LSTM dan CNN BiLSTM dengan fitur 50 dan 7 data historis. Hasil penelitian menunjukkan bahwa penambahan jumlah lapisan memiliki korelasi dengan performa prediksi harga penutupan saham. Model CNN BiLSTM dengan fitur 50 data historis merupakan model yang terbaik dalam memprediksi harga saham dengan nilai MSE, MAE, dan RMSE secara berturut-turut adalah 0.12481, 0.29218, dan 0.3520707 pada lapisan kelima saham ITMG, diikuti oleh model CNN LSTM dengan fitur 50 data historis dengan nilai MSE, MAE, dan RMSE secara berturut-turut adalah 0.02965, 0.13582, dan 0.171661 pada lapisan kelima saham ADRO, CNN BiLSTM dengan fitur 7 data historis dengan nilai MSE, MAE, dan RMSE secara berturut turut adalah 0.04988, 0.20037, dan 0.2232269 pada lapisan ketiga saham ASII, dan CNN LSTM dengan fitur 7 data historis dengan nilai MSE, MAE, dan RMSE secara berturut-turut adalah 0.03705, 0.15974, dan 0.1924773 pada lapisan kelima saham ADRO. / Technological advances make it possible to predict stock prices using machine learning methods. Deep learning methods that will be used are Convolutional Neural Network Long Short Term Memory and Convolutional Neural Network Bidirectional Long Short Term Memory. The purpose of this research is to determine the impact of the number of layers on the model and use evaluation metrics to determine which model is better at predicting stock prices. The models are CNN LSTM and CNN BiLSTM models featuring 50 and 7 historical data. The results showed that the addition of the number of layers has a correlation with the predicted performance of the stock's closing price. The CNN BiLSTM model featuring 50 historical data is the best model in predicting stock prices with MSE, MAE, and RMSE values respectively are 0.12481, 0.29218, and 0.3520707 in the fifth layer of ITMG stocks, followed by the CNN LSTM model featuring 50 historical data with MSE, MAE, and RMSE values respectively are 0.02965, 0.13582, and 0.171661 on the fifth layer of ADRO stock, CNN BiLSTM featuring 7 historical data with MSE, MAE, and RMSE values respectively are 0.04988, 0.20037, and 0.2232269 on the third layer of ASII stocks, and CNN LSTM featuring 7 historical data with MSE, MAE, and RMSE values respectively are 0.03705, 0.15974, and 0.1924773 on the fifth layer of ADRO shares.

Item Type: Thesis (Bachelor)
Creators:
CreatorsNIMEmail
Widjaya, JeffreyNIM01112190022widjaya.jeffrey@gmail.com
Contributors:
ContributionContributorsNIDN/NIDKEmail
Thesis advisorSaputra, Kie Van IvankyNIDN0401038203kie.saputra@uph.edu
Thesis advisorKrisnadi, DionNIDN0316029002dion.krisnadi@uph.edu
Uncontrolled Keywords: cnn, lstm, 50 data historis, 7 data historis, hasil, analisis
Subjects: Q Science > QA Mathematics
Divisions: University Subject > Current > Faculty/School - UPH Karawaci > Faculty of Science and Technology > Mathematics
Current > Faculty/School - UPH Karawaci > Faculty of Science and Technology > Mathematics
Depositing User: Jeffrey Widjaya
Date Deposited: 24 Jan 2023 03:57
Last Modified: 24 Jan 2023 03:57
URI: http://repository.uph.edu/id/eprint/52846

Actions (login required)

View Item View Item