- Institute of Graduate Studies - department Big Data Analytics - Thesis (Master) - Includes references (47-50 p.) - This thesis examines the use of LSTM neural networks to forecast the closing price of the NASDAQ index using historical price data. The study makes use of a dataset that spans many years and includes daily closing prices, volume, and other NASDAQ index elements. LSTM networks are used to forecast future closing prices based on historical data, with an emphasis on a 60-day prediction window. Metrics like R2 , Mean Absolute Percentage Error (MAPE), and Root Mean Squared Error (RMSE) are used for evaluating the performance of the LSTM model. The findings show that the LSTM model predicts future closing prices well, with lower error rates than baseline models. The research additionally examines at how different input parameters and network structure influence the model's performance. Overall, the study indicates that LSTM networks are beneficial in predicting the price of index and explain their potential for application in financial forecasting applications. - Text in English.
University of Kyrenia Grand Library1st Floor (HG4515.5 .A35 2023)
Reference Section
9663807771
Item available
University of Kyrenia Grand LibraryGrnd. Floor (HG4515.5 .A35 2023)
Reference Section
NEAR EAST UNIVERSITY GRAND LIBRARY +90 (392) 223 64 64 Ext:5536. Near East Boulevard, Nicosia, TRNC This software is developed by NEU Library and it is based on Koha OSS
conforms to MARC21 library data transfer rules.