Abstract
The digital age continues to advance, resulting in an increase in the volume, variety and of velocity (3Vs) data. This has significant implications for the decision-making processes within organizations. As advanced analytical techniques are being used by different types of industries, including healthcare, finance, manufacturing and retail. They have become a key driver of innovation and competitive advantage. This study applies a systematic review methodology to summarize and evaluate the most commonly used analytical techniques for informed decision-making. This study analyses the applications, effectiveness and skill sets needed for these techniques. It also maps key trends and sector-specific implementations from 27 peer reviewed literature published in Scopus from 2019 to 2024. It relies on a PRISMA-guided systematic review, which combines bibliometric mapping with VOS viewer, text mining and thematic analysis to identify leading techniques and its sectoral applications.
The key findings indicates that there has been significant transformation shift from descriptive analytics toward predictive and prescriptive analytics, resulting from machine learning (ML) and artificial intelligence (AI) that are advanced. The use of techniques such as time series forecasting, deep learning, random forests and support vector machines is increasing in order to uncover operational inefficiencies, optimize forecasting accuracy and uncover complex patterns. The study also identifies the importance of business intelligence (BI) in providing access to data-driven insights across organizational levels using dashboarding, data visualization and self-service analytics. The study also reveals that deep learning models have a strong adoption in the healthcare sector with a high diagnostic accuracy. This is followed by the finance, retail, energy and manufacturing sectors for their fraud detection, demand forecasting, resource optimisation and predictive maintenance respectively.
Although these advancements have been made, methodological and organizational challenges still exist. Analytics techniques are still not fully utilised due to issues of model interpretability, integration and data quality, ethical governance and the lack of professionals with technical and domain knowledge. To foster trust and adoption for the identified challenges, it is crucial to establish robust data governance frameworks and implement continuous professional development and explanations of AI models. In conclusion, the study emphasises the need for context-sensitive evaluation of...