mrEnsemble is an environment for Distributed Machine Learning, which allows easy creation and deployment for ensembles of arbitraty methods. The goal is to have a handy tool to analyse the interaction and performance of aggregated statistical models.
The Framework is based on the Hadoop infrastructure, while the models are being trained using MapReduce. The project is written in Python with MrJob.
Work was supported by AWS in Education Grant award.