Multi-Step Neutrosophic Cognitive Map Based Decision Making Framework for Short-Term Financial Stock Market Price Trend Prediction

Neutrosophic cognitive maps are expansion of fuzzy cognitive maps, containing indetermination in causal relations. Fuzzy cognitive maps do not require an indeterminate relationship, making it less adequate for real-time applications. A logic in which every proposition is projected to have the truth percentage in subset T and the falsity percentage in subset F is named Neutrosophic Logic. This logic is also considered the general form of Intuitionistic fuzzy logic. Stock price prediction is a main topic in economics and finance, which has promoted the priority of investigators in recent years to improve improved predictive methods. Predicting price and tendency of the stock market denote indispensable features of finance and investment. Many scientists have presented their ideas to predict the market price to make money while trading utilizing different methods like statistical and technical analysis. This manuscript proposes a Neutrosophic Cognitive Map-Based Short-Term Financial Stock Market Price Trend Prediction (NCM-SFSMPTP) model. The main goal of NCM-SFSMPTP technique relies on improving the accurate approach for stock market price trend prediction. At first, the min-max normalization methodology is utilized in the data normalization phase to standardize and scale data for consistency, comparability, and efficient processing. For the classification process, the neutrosophic cognitive map (NCM) technique is employed. Finally, the improved arithmetic optimization algorithm (IAOA)-based hyper-parameter selection is implemented to enhance the classification outcomes of the NCM system. The performance validation of the NCM-SFSMPTP methodology is verified under the Apple Stock Price Trend and Indicators dataset and the outcomes are determined regarding to several measures. The experimental validation of the NCM-SFSMPTP method illustrated a superior accuracy value of 94.79% over existing models in stock market price trend prediction process. © 2025, American Scientific Publishing Group (ASPG). All rights reserved.