The Federation – A novel machine learning technique applied to data from the Higgs Boson Machine Learning Challenge

Abstract The Federation is a new machine learning technique for handling large amounts of data in a typical high energy physics analysis. It utilizes UMAP to create an initial lowdimensional representation of a given set of trainings data. The dataset in this representation is clustered by using HDBSCAN. These clusters can then be used for a federated learning approach, in which we separately train one classifier for each cluster on the high-dimensional data. As a requirement for this approach, we need to apply an Imbalanced Learning method [9] to the data in the found clusters before the training. By using a Dynamic Classifier Selection method, the Federation can then make predictions for the whole dataset. As a proof of concept for this novel technique, open data from the Higgs Boson Machine Learning Challenge [1] is used and comparisons to results from established methods will be presented.