Prescriptive analytics: a survey of emerging trends and technologies

This paper provides a survey of the state-of-the-art and future directions of one of the most important emerging technologies within business analytics (BA), namely prescriptive analytics (PSA). BA focuses on data-driven decision-making and consists of three phases: descriptive, predictive, and prescriptive analytics. While descriptive and predictive analytics allow us to analyze past and predict future events, respectively, these activities do not provide any direct support for decision-making. Here, PSA fills the gap between data and decisions. We have observed an increasing interest for in-DBMS PSA systems in both research and industry. Thus, this paper aims to provide a foundation for PSA as a separate field of study. To do this, we first describe the different phases of BA. We then survey classical analytics systems and identify their main limitations for supporting PSA, based on which we introduce the criteria and methodology used in our analysis. We next survey, categorize, and discuss the state-of-the-art within emerging, so-called PSA+\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$^+$$\end{document}, systems, followed by a presentation of the main challenges and opportunities for next-generation PSA systems. Finally, the main findings are discussed and directions for future research are outlined.