Business Decision Support Systems Based On Deep Learning And Cognitive Computing

Abstract

The purpose of this paper is to explore the use of deep learning and cognitive computing in order to support decision-making in dynamic and complex situations in a business environment using Business Decision Support Systems. Conventionally, DSS were rule-based systems, which might be constrained by limited structured data sets and were used for offline decisions only. This paper aims to propose a Hybrid DSS architecture incorporating deep learning techniques such as CNNs, RNNs, and LSTMs, as well as cognitive computing techniques like NLP and Knowledge Graph. The method involves developing deep learning algorithms on the basis of structured data while employing cognitive computing to make inferences from unstructured data. The proposed system will be tested based on metrics of accuracy, precision, recall, F1 measure, RMSE, and AUC-ROC. As a result, the Proposed Hybrid DSS performs better than Rule-Based DSS with the accuracy of 94.2%, precision of 93.5%, F1 of 93.1%, RMSE of 17.6%, and AUC-ROC of 96%, where Rule-Based DSS has accuracy of 78.4% and ML DSS – 87.6%. The findings confirm the efficacy of the hybrid system in enhancing the application of deep learning and cognitive computing in the process of prediction, understanding, and reasoning. This study proves that the suggested hybrid decision support system framework offers distinct benefits compared to traditional decision support system frameworks, making it a more adaptable, precise, and context-aware decision-making system. Further research is being conducted on computational efficiency and real-time processing concerns.