http://digilib.uin-suka.ac.id/id/eprint/76841 2026-06-19T08:00:58Z info:eu-repo/semantics/descriptiveMetadata https://digilib.uin-suka.ac.id/id/eprint/76841/ PERBANDINGAN MODEL LONG SHORT-TERM MEMORY (LSTM) DAN GATED RECURRENT UNIT (GRU) DALAM PERAMALAN HARGA SAHAM (STUDI KASUS : INDEKS HARGA SAHAM GABUNGAN (IHSG) PERIODE JANUARI 2010 - DESEMBER 2025) Irodatul Jannah, NIM.: 22106010047 515.6 Metode Analitik - Matematika This study aims to compare the effectiveness of RNNbased neural network architectures, namely LSTM and GRU, in forecasting the weekly closing prices of the IHSG from January 2010 to December 2025. Four model architectures were employed, namely LSTM–LSTM, GRU–GRU, LSTM– GRU, and GRU–LSTM. The research stages included data normalization, sliding window transformation, and splitting the data into training and testing sets. All models were trained using the Adam optimizer with the same hyperparameter configuration. Model performance was evaluated using MSE and MAPE. The results showed that the GRU– GRU architecture achieved the best performance with the lowest MSE and MAPE values, indicating that it was more effective in capturing the time series patterns of IHSG closing prices. 2026-04-30 Thesis NonPeerReviewed text id https://digilib.uin-suka.ac.id/id/eprint/76841/1/22106010047_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf text id https://digilib.uin-suka.ac.id/id/eprint/76841/2/22106010047_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf Irodatul Jannah, NIM.: 22106010047 (2026) PERBANDINGAN MODEL LONG SHORT-TERM MEMORY (LSTM) DAN GATED RECURRENT UNIT (GRU) DALAM PERAMALAN HARGA SAHAM (STUDI KASUS : INDEKS HARGA SAHAM GABUNGAN (IHSG) PERIODE JANUARI 2010 - DESEMBER 2025). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA. info:eu-repo/semantics/objectFile info:eu-repo/semantics/objectFile info:eu-repo/semantics/humanStartPage