Pentapartitioned Neutrosophic Vague Soft Set with Optimization Algorithm Based Business Intelligence Framework for Data-Driven Demand Forecasting Model

Neutrosophic logic is a neonate research field in which all propositions are anticipated to have the percentage (proportion) of truth in a sub-set T, the proportion of falsity in a sub-set F, and the proportion of indeterminacy in a sub-set I. Neutrosophic set (NS) is efficiently applied for indeterminate information processing and provides assistance to address the indeterminacy information of data. Demand Forecasting, undoubtedly, is the only most significant element of some organization's Supply Chain. It defines the predictable demand for the future and sets the preparedness level that is needed on the supply side to match the demand. Business intelligence (BI) plays a significant part in helping the decision maker obtain the understanding for increasing productivity or improved and faster decisions. Furthermore, it improves and helps the efficacy of functional rules and its influence on corporate-level decision-making that provides improved strategic options in dynamic business environments. Within the period of data-driven demand forecasting, the integration of artificial intelligence (AI) technologies in BI models has transformed the system groups that utilize and analyze data. In the manuscript, a Business Intelligence Framework for a Data-Driven Demand Forecasting Model Using a Pentapartitioned Neutrosophic Vague Soft Set (BIFDDF-PNVSS) technique is proposed. The main goal of the BIFDDF-PNVSS technique is to progress the accurate BI structure for the demand forecasting method. The data pre-processing stage is initially applied for converting input data into a beneficial format by the Z-score normalization method. Moreover, the PNVSS technique is utilized for the data-driven demand prediction model. Finally, to improve the prediction performance of the PNVSS model, the parameter tuning process is performed by implementing the cheetah optimization algorithm (COA) model. A comprehensive experimentation is performed to verify the performance of the BIFDDF-PNVSS methodology under the demand forecasting dataset. The BIFDDF-PNVSS methodology outperforms existing techniques with a superior MSE of 0.0008, demonstrating its exceptional accuracy in demand forecasting compared to other models.