The OpenML 2017 Workshop will be 9-14 October in Leiden, the Netherlands. Save the date!
Dive into the useR!2017 OpenML Tutorial.
Use your favorite OpenML tool to participate in the (external) Open Algorithm Selection Competition 2017. Deadline is August 27, 2017.
What if you could explore machine learning research as easily as exploring Wikipedia?
What if you could share new data, code and experiments as easily as sending a tweet?
OpenML enables truly collaborative machine learning. Scientists can post important data, inviting anyone to help analyze it. OpenML structures and organizes all results online to show the state of the art and push progress.
OpenML is being integrated in most popular machine learning environments, so you can automatically upload all your data, code, and experiments. And if you develop new tools, there's an API for that, plus people to help you.
OpenML allows you to search, compare, visualize, analyze and download all combined results online. Explore the state of the art, improve it, build on it, ask questions and start discussions.
What if you could collaborate on hard problems with hundreds of scientists at once?
What if you could easily access the latest data to answer questions or verify findings?
Science follows certain methods. On OpenML, these are expressed as tasks detailing what results must be uploaded, and requiring the information necessary to ensure that uploaded results are interpretable and verifiable.
Tasks are solved collaboratively. Anyone can propose new tasks, and anyone can upload new results, augment the data, contribute new ideas, ask questions, or discuss issues and results online.
OpenML enriches research output by making it freely accessible, organized, continuously updated, immensely detailed, and reproducible. It stimulates online discussion and diminishes publication bias.
What if you could spend more time doing actual research?
What if you could get more credit for your work by making it more visible?
OpenML takes care of the routinizable work required for yielding insight. It helps you run large amounts of experiments using many datasets and techniques, organizes them online and relates them to the state-of-the-art.
OpenML organizes, links and annotates your results so that you can focus on discovery: interpreting the data, linking it to other data, formulating new hypotheses and designing new experiments to test them.
OpenML helps you share your results for optimal impact. Share with the world or with friends, freely or with attribution (citation). Follow how often your contributions are reused or liked and climb up the leaderboards.