Analisis Harga Saham LQ45 dengan Metode Cluster Time Series dan Hybrid ARIMA-ANN

Apriyanto, Fajar (2023) Analisis Harga Saham LQ45 dengan Metode Cluster Time Series dan Hybrid ARIMA-ANN. Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 06211940000107-Undergraduate_Thesis.pdf] Text
06211940000107-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 October 2025.

Download (3MB) | Request a copy

Abstract

Saham merupakan surat berharga yang menunjukkan kepemilikan modal dalam suatu perusahaan yang harganya sangat fluktuatif menjadikannya memerlukan analisis yang dapat mengelompokkan berdasarkan harga sahamnya serta memprediksi pergerakan harganya di masa mendatang. Pada penelitian ini dilakukan pengelompokan Saham LQ45 menggunakan data harga penutupan harian menggunakan metode cluster time series yaitu metode pengelompokan pada data runtun waktu berdasarkan data historisnya. Penelitian ini menggunakan metode complete linkage dengan jarak dynamic time warping yang cukup baik untuk diterapkan pada analisis cluster time series. Selanjutnya analisis dilanjutkan dengan peramalan harga Saham LQ45 untuk mengetahui prediksi harganya berdasarkan cluster yang telah terbentuk menggunakan metode hybrid ARIMA-ANN. Penggabungan kedua metode ini akan menggunakan karakteristik unik dan kekuatan masing-masing dari model. Berdasarkan hasil analisis, Saham LQ45 dapat dikelompokkan ke dalam dua kelompok dengan nilai Silhouette sebesar 0,883 yang termasuk dalam strong cluster. Cluster kedua terdiri dari dua saham dengan rata-rata harga lebih tinggi daripada cluster pertama yang terdiri dari 42 saham. Peramalan dengan metode hybrid ARIMA-ANN lebih sesuai diterapkan pada saham INTP sedangkan metode tunggal ARIMA lebih sesuai diterapkan pada saham ITMG. Kedua model tersebut menghasilkan error MAPE di bawah 1,5% dan di bawah 2,1% dalam meramalkan saham INTP dan ITMG.
===============================================================================================================================
Stocks are securities that show capital ownership in a company whose price fluctuates very much, making it require analysis that can classify based on share prices and predict future price movements. In this study, the grouping of LQ45 stock is carried out using daily closing price data using the cluster time series method, namely the grouping method on time series data based on historical data. This study uses the complete linkage method with dynamic time warping distance which is good enough to be applied to cluster time series analysis. The next analysis is followed by forecasting the LQ45 stock price to predict its price based on the clusters that have been formed using the hybrid ARIMA-ANN method. Combining these two methods will take advantage of the unique features and strengths of each model. Based on the research, LQ45 stock can be grouped into two groups with a Silhouette score of 0.883 which is included in the strong cluster. The second cluster consists of two stocks with an average price higher than the first cluster which consists of forty two stocks. Forecasting with hybrid ARIMA-ANN method is more suitable to be applied to INTP stock while single method of ARIMA is more suitable to be applied to ITMG stock. Both models produce the error of MAPE below 1.5% and below 2.1% in forecasting INTP and ITMG stocks.

Item Type: Thesis (Other)
Uncontrolled Keywords: Cluster Time Series, Dynamic Time Warping, Hybrid ARIMA-ANN, LQ45, Silhouette Score
Subjects: H Social Sciences > HA Statistics > HA30.3 Time-series analysis
H Social Sciences > HG Finance > HG4529 Investment analysis
H Social Sciences > HG Finance > HG4910 Investments
Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry)
Q Science > QA Mathematics > QA278.55 Cluster analysis
Q Science > QA Mathematics > QA280 Box-Jenkins forecasting
Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Q Science > QA Mathematics > QA278 Cluster Analysis. Multivariate analysis. Correspondence analysis (Statistics)
Divisions: Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Fajar Apriyanto
Date Deposited: 06 Sep 2023 01:36
Last Modified: 06 Sep 2023 01:36
URI: http://repository.its.ac.id/id/eprint/104437

Actions (login required)

View Item View Item