Applied Statistics with Single-Valued Neutrosophic Fuzzy Soft Expert Sets for Market Trend Forecasting Model

Mohanned H. Alharbi¹, Khalil A. Alruwaitee², Sulima M. Awad Yousif³, Ashraf A. Awad Alotaibi⁴, Abdelgalal O. I. Abaker^5,*, Azhari A. Elhag⁶

¹Finance and Business Sector, Institute of Public Administration, Riyadh 11141, P. O. Box 205, Saudi Arabia;

²Department of Economics and Finance, Business College, Taif University, P. O Box 11099, Taif 21944, Saudi Arabia;

³Department of Basic Science, College of Preparatory year, Najran University, Najran, 61441, Saudi Arabia

⁴Department of Psychology, College of Education, King Khalid University, Abha, Saudi Arabia

⁵Applied College, Khamis Mushait, King Khalid University, Abha, Saudi Arabia

⁶Department of Mathematics and Statistics, College of Science, P.O. Box 11099, Taif University, Taif 21944, Saudi Arabia

Email: Harbimh@ipa.edu.sa; K.gamith@tu.edu.sa; smyoussef@nu.edu.sa; asalotaibi@kku.edu.sa; aoadrees@kku.edu.sa; a.alhag@tu.edu.sa

Abstract

Applied statistics has been instrumental in predicting behaviours and future market trends. In the field of financial time series analysis, the incorporation of deep learning (DL) methods and applied statistics has made a significant contribution to the prediction model. Practitioners and researchers can extract complex features and dependencies from past financial data by leveraging neural network structures like long short-term memory (LSTM) and recurrent neural networks (RNNs). These DL approaches advance the development of predictive models prone to forecasting different financial metrics, such as asset returns, stock prices, and market volatility, with outstanding accuracy. With the combination of statistical approaches with DL techniques, researchers can leverage the power of both worlds to make more informed investment decisions and improve forecasting capabilities in volatile and dynamic financial markets. This study develops a new Applied Statistics with Single Valued Neutrosophic Fuzzy Soft Expert Sets (AS-SVNFSES) technique for Financial Time Series Forecasting. The presented AS-SVNFSES technique aims to forecast the input financial time series data. The AS-SVNFSES technique primarily applies data preprocessing using a Z-score normalization approach. For the forecasting of financial data, the AS-SVNFSES technique makes use of the SVNFSES technique. Finally, the parameter tuning of the SVNFSES technique is performed using the chimp optimization algorithm's (ChOA) design. A series of experimentations have illustrated the amended performance of the AS-SVNFSES model. The experimental value inferred that the AS-SVNFSES technique gains improved performance over other models.

Keywords: Neutrosophic Set; Financial Time Series; Chimp Optimization Algorithm; Data Preprocessing; Neutrosophic Soft Sets