scikit-learn

PyData Paris 2024 45-minute Keynote by Olivier Grisel: ?? Handling predictive uncertainty in machine learning Machine Learning practitioners build predictive models from "noisy" data resulting in uncertain predictions. But what does "noise" mean in a machine learning context? #python #machinelearning #datascience #opensource https://lnkd.in/ez2hn8dp

Are you one of the 1.5 billion (yes, BILLION) people who’ve downloaded scikit-learn? ??? (eta: scikit-learn has been downloaded 1.5 billion times. Likely you've downloaded it more than once; still--a phenomenal number! ?? ) A cornerstone for machine learning in Python, scikit-learn has become an essential tool for data scientists worldwide. While I knew it was an open-source library, I was surprised to learn about its French origins and the pivotal support it received from Inria (the National Institute for Research in Digital Science and Technology) and the French government. ???? Curious to learn more? ?? Check out the link in the comments to an interesting Practical AI interview with Yann Lechelle, CEO of the spin-out Probabl that now supports sci-kit learn, and one of its core developers, Guillaume Lemaitre. Fun facts I learned : ? Despite the rise of deep learning and GenAI, scikit-learn’s foundational models still account for 95%+ of current machine learning use cases. ? 22% of scikit-learn downloads come from the U.S. alone. ? Behind the library are 10 full-time developers, supported by a passionate community of volunteer contributors who keep it thriving. If you’re a fan of scikit-learn (or just curious about its impact), give the interview a listen!

Did you know that many #opensource repositories suffer from a large number of open pull requests that need a reviewer? Open source maintainers are usually very busy with many of them being volunteers. This easily leads to a great number of pull requests that lack attention. So this Hacktoberfest I've been focussing on reviewing pull request and I encourage you to do the same! ?? Many people are afraid of starting to review code because they fear that they lack the required knowledge and experience. However, if you have made a few contributions of a similar kind to a project, for example code documentation, you definitely have the necessary skills to start reviewing code documentation! Starting to review open source code is very educational, increases your competence over time and will make you feel more connected to the community of the project you are involved with. For more information about the different aspects of reviewing a pull request, check out this video: https://lnkd.in/eQ2mEAQ5

Tick-tock, tick-tock ? The user survey is closing soon... Participate to this collaborative effort to improve your favorite package! We would like to thank the teams from University of Oxford, POSSEE OpenTeams, and :probabl. as well as many scikit-learn contributors, for their time and effort in designing and translating it. https://lnkd.in/eYNG2bjd

?? Bugfix release - scikit-learn 1.5.2 is out! ?? 6 fixes It fixes several regressions introduced in version 1.5 More details in the changelog: https://lnkd.in/emPUn6Rt You can upgrade with pip as usual: pip install -U scikit-learn or using the conda-forge builds: conda install -c conda-forge scikit-learn Thanks to all the contributors!

Yay, the 2024 scikit-learn user survey is out! Please join this structured dialogue with the scikit-learn team to better guide and prioritize decision-making about the development of the project. ?? We have the survey available?in these languages: Arabic, English, French, Japanese, Mandarin, Portuguese, Spanish. ?? The survey will take about 15 minutes of your time and close on October 14th, 2024. ?? A survey results report and anonymized dataset will be publicly released by the end of 2024. ?? To get started, click here: https://lnkd.in/eyAyjEu2

Thrilled to share that scikit-learn has been awarded a CZI-EOSS grant for cycle 6! ?? With this funding, we'll enhance and expand tools for predictive model evaluation and inspection. Read all about it here: https://lnkd.in/eCpr6uji ?? A big thank you to Chan Zuckerberg Initiative and Wellcome Trust for their support! #MachineLearning #OpenScience #python #DataScience

?? There is now a handy widget on the scikit-learn GitHub repo for citations. ??GitHub has added this built-in citation support so researchers and scientists can more easily receive acknowledgments for their contributions to software. ??Check it out! If you use scikit-learn in a scientific publication, we appreciate citations. https://lnkd.in/dwncfBb7 #opensource #datascience #machinelearning

I added an enhancement to scikit-learn! If you have used the "sklearn.utils.validation.check_array" function to prep your data before passing it to a model, you know how convenient it is. It is not even more by checking that your array doesn't contain negative values. The proposed changes also sparked additional API changes (see both entries here: https://lnkd.in/eDK7zxkw). The improvement idea came by noticing a pattern that both the scikit-learn repo itself and the other users across GitHub had - using the function "check_non_negative" right after "check_array" in the cases where they would like to make sure that all values are non-negative. I encountered this while working on making fairlearn's estimators more scikit-learn compatible. If you are creating custom scikit-learn estimators or looking into improving existing ones, enabling the new v0.15 estimator checks (https://lnkd.in/ebhFatU7) could be a great way of giving you a structured way to achieve this (read more here: https://lnkd.in/e-jsB7HA). I've been slowly chipping in there, which has led to lots of refactoring, testing improvements, and overall safer code. Lastly, check out fairlearn and all the contributions in the last few months, which make it easier to build, contribute and use: https://lnkd.in/ez-dmKvn I've been a part of :probabl. since May and it's been the most fun I have had at work in years. I plan to have a series of technical talks about the work we are doing. I can't wait to tell you all about it :)

?? Please welcome Adam Li, PhD to the scikit-learn team. Adam is currently a Postdoctoral Research Scientist at Columbia University in the Causal Artificial Intelligence Lab. Learn about his path through open source to scikit-learn as well as his current research: https://lnkd.in/e_Q7DXiz