What is prescriptive analytics? When reading about 'Predictive to Prescriptive' analytics, I am not able to find good examples of what it actually means. 
I know that a predictive model can only tell me what will happen. And in real business scenarios this is not enough. I need to know what can I do to handle that predicted thing for better results and that's were prescriptive stuff comes in. And I read that optimization is a major part of this step.
Therefore, my curiosity is how to do prescriptive analytics? Any examples? something like Kaggle(though I searched)? How a randomForest/logistic reg model is converted for decision making through optimization ?
 A: Q: How to do prescriptive analytics? A: Do optimization and call it a new buzzword, prescriptive analytics.
Example: Use predictive analytics to forecast sales of various products at various prices. Feed that into an optimization to determine optimal pricing across the product portfolio, inventory levels, etc.
A: Firstly we have descriptive analytics such as reporting what daily sales were over some period of time. Then we might have exploratory analytics which might detect that Sunday sales are statistically significantly decreasing at a higher rate than any other day and when this trend started.  These inferential analytical results might suggest that one might address oneself to ferreting out possible causes for this phenomenon. To make future realizations for sunday sales different than the expected i.e. downwards trending, cost-effective future actions might be considered . 
To be able to change the future (predictive analytics) from what is most likely to happen one needs to identify the significant factors underlying/driving the data we observe.
