Open Machine Learning
OpenML | An Open Project Spotlight
Heidi Seibold is a biostatistics PhD student at the University of Zurich, an OpenML core member, and a big fan of open science. Heidi brought OpenML to our current round of Mozilla Open Leaders where she’s been improving the experience for non-technical users.
I interviewed Heidi to learn more about OpenML and how you can help!
What is OpenML?
OpenML stands for Open Machine Learning and is an open science, open source and open data project. The OpenML online platform allows people to share and discover data, questions, machine learning algorithms and results. We offer open source tools to download everything you want to your favourite machine learning environments and work with it. You can then upload your results back onto the platform so that others can learn from you. If you have data, you can use OpenML to get insights on what machine learning method works well to answer your question. Machine Learners can use OpenML to find interesting data sets and questions that are relevant for others and also for machine learning research (e.g. learning how algorithms behave on different types of data sets).
OpenML is an ongoing project and it is far from perfect. But IMHO it is the best of its kind and you can already do very interesting stuff with it. Also, since it is an open source project, everyone can help making it even better 😎
What part of OpenML have you been working on during the mentorship program?
Mostly, making OpenML more attractive to people who have data but no or little machine learning experience. Machine learners find the project exciting right away, but for applied scientists and other data owners it is difficult to understand the benefits it might have for them. Other goals were to start a foundation that supports the OpenML project and making it easier for people to get involved.
What are you most proud of accomplishing during Hacktoberfest?
This may sound small, but it was exciting to realize that if you make it easy for people to do tasks, they will appear out of nowhere and help you. I think it will help me in the future.
What challenges have you faced working on this project?
OpenML has grown into quite a complex project. It is very hard to document and organize such a big project in a way that new people can come in and eventually help with more complex things. Especially since OpenML is a side or hobby project for all of us and guiding new people into the project costs a lot of time.
How has your project been impacted by Mozilla Open Leaders?
Mozilla Open Leaders has made me think about many things and we discussed many of the ideas with the OpenML team. I think there are many small things that came out of this: for example we now decided to have issue tags on GitHub inviting new people to help, we are very close to actually starting a foundation now and we have been working on little videos explaining machine learning and OpenML.
How can others help you continue the work on OpenML?
This one is easy, since we have a dedicated file for this now. It is one of the things we created because of Hacktoberfest. There is so much to be done, also for people who do not know how to program at all: Check out the website and let us know what could be improved; Try OpenML and talk about your experience on your blog to help us get more widely known and learn from problems you experienced; Help us documenting the python package that connects to OpenML. Check out our CONTRIBUTING.md if you are interested in helping or simply contact us!