Artificial intelligence beats human experts

    20.11.2015 10:31
    Repeated success in the Automatic Machine Learning Challenge (AutoML)

    A collaborative team of Dr. Frank Hutter's Automated Algorithm Design group and Dr. Joschka Boedeker's Machine Learning group succeeded again in the second phase of the Automatic Machine Learning (AutoML) Challenge. The challenge aims at the development of automatic systems for machine learning on datasets of all kinds (e.g., medical diagnosis, speech recognition, and object recognition). The study of such data sets by human machine learning experts is traditionally a lengthy and manual process. This competition now aims to automate this process and to make good predictions as quickly as possible.

    Based on its automated methods, the IIF group won the 1st place in the competition against 125 teams of human experts and the 2nd place in a fully-automated contest, thus winning $2500 of prize money. Machine learning is rapidly gaining importance in today's information society and the demand for experts far exceeds the capacity of university programs. The fact that the artificial intelligence developed by the IIF team now for the first time won a competition against such human experts may thus have far-reaching implications.

    The IIF team will present a scientific article about its methods at the world-leading machine learning conference Neural Information Processing Systems December 7-12 in Montreal, Canada.

    A preprint of that article can be found here: http://aad.informatik.uni-freiburg.de/papers/15-NIPS-auto-sklearn-preprint.pdf

    The team's system is available online and allows even machine learning novices to use it effectively: https://github.com/automl/auto-sklearn

    More information about the AutoML Challenge can be found here: http://automl.chalearn.org/



    Benutzerspezifische Werkzeuge