Bridging the gap between predictive and prescriptive analytics - new optimization methodology needed

Business analytics is becoming more and more important nowadays. Up to now predictive analytics appears to be much more applied in practice than prescriptive analytics. We argue that although optimization is used to obtain predictive models, and predictive tools are used to forecast parameters in optimization models, still the deep relation between the predictive and prescriptive analytics is neither well understood nor fully exploited. We describe two opportunities to really exploit the synergy between the predictive and prescriptive part. The first is to perform optimization by directly using the predictive models. Adding optimization functionality in predictive analytics tools could be of huge added value for practice. The second opportunity is to replace manual model building with automated data-driven model building, using modern predictive analytics. The pros and cons for such a way of optimization are also discussed.