Talk:Leveraging self-learning models to ensure optimization in new situation

From PegaWiki
Jump to navigation Jump to search

When describing how to interpret the bubble chart, the x-axis is being described as to what extent the proposition is relevant to the customer. This is not correct, as the x-axis measures the model performance. The y-axis depicts overall accept rate so that could be seen as a measure of relevance - though 'accepted often' is a more to the point description. The x-axis simply stands for how well the model can differentiate between successful and unsuccessful decisions - in this example differentiate between customers who are more or less likely to respond positively to a proposition.