Multipatch predictive selection
Abstract
Predictive selection, with all its variants, is one of the most powerful of all the meta-algorithmic patterns. However, it is more than a single pattern. In this chapter, the use of predictive selection as a pattern for both better predictive systems and for incorporating individual predictive systems into a larger predictive system is shown. Improving predictive power by partitioning the input space is shown to be a trade-off between high accuracy and overtraining. Finally, multipath predictive selection is shown to be a predictive approach that has similarities to decision trees. The reader is left with a repertoire of approaches to predict the optimal system deployment for a wide range ...
Get Meta-Analytics now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
