Building and using stacked machine learning models –  A proven path to more accurate models – IBM Developer

Build cloud-native applications for regulated workloads with IBM Cloud for Financial Services Learn more

Building and using stacked machine learning models –  A proven path to more accurate models

Stacking is the process of using the predictions of multiple machine learning models as the inputs of another model. When done correctly it can minimize the bias of the individual input models with respect to the training data and produce a an aggregated model that does a better job of generalizing to unseen data. For this reason it is aa very popular technique among data science competition winners. In this session, geared towards developers, data scientists and data analysts, you’ll learn how to build a stacked ensemble model in Python and some proven tips and techniques that will allow you to add this proven technique to your data science toolbox.