Machine Learning for Business Analytics: Enhancing Forecasting Accuracy

Abstract

Business intelligence (BI) has shifted significantly as a result of the rapid advancement of digital technology, opening up new opportunities for strategic insights and data-driven decision-making. This study explores the relationship between these two fields, concentrating on how information management systems and artificial intelligence (AI) may work together to improve financial predictions' precision, dependability, and scalability. A key component of successful predictive modelling is data warehousing, which unifies enormous volumes of historical and current financial data from several sources into a single repository. With the use of this aggregated data, AI-powered prediction models—which include deep learning architectures, machine learning algorithms, and other sophisticated statistical methods—can provide accurate and useful forecasts and nuanced insights. With a thorough assessment of 152 publications from 1969 to 2023, this study methodically analyses and compares cutting-edge supply chain (SC) forecast techniques and technologies throughout a certain time period. In order to forecast the effects on the human workforce, inventory, and SC as a whole, a novel framework that incorporates Big Data Analytics into SC Management (problem identity, data sources, and exploratory analysis of information, machine-learning model training, hyperparameter adjustment, performance evaluation, and optimisation) has been proposed. First, the need of gathering data in accordance with the SC strategy and the methods for doing so have been covered.