Scaling scikit-learn - Introducing New Sets of Computational Routines

Explore the latest advancements in scaling scikit-learn's performance in this conference talk from EuroPython 2022. Delve into the efforts of core developers to enhance the computational efficiency of this widely-used open-source machine learning library. Gain insights into the PyData ecosystem and CPython interpreter's impact on performance, identify computationally expensive patterns, and understand the technical decisions behind new routine implementations. Learn about the project's approach to balancing efficiency with existing requirements and get a glimpse of future developments in hardware-specialized computational routines. Discover how scikit-learn is evolving to meet the demands of modern machine learning applications while maintaining its position as a leading scientific library for data analysis.

Scaling scikit-learn - presented by Julien Jerphanion