Comparación de H2O y MLib como herramientas para el aprendizaje automático en plataformas de Internet de las Cosas

Abstract

In this work, two machine learning tools, H2O and MLib, were compared with the objective of selecting one of them to be integrated into an Internet of Things platform. For the comparison, the execution time of the tools was taken into account while executing the K-means algorithm on the same data set. In this test, H2O showed the best results, and this added to its easy usability and ability to integrate with other machine learning tools make it stand out on the wide variety of options that exist for machine learning.