<div dir="ltr"><div style="font-size:small" class="gmail_default">We're happy to announce the 0.23.1 release which fixes a few issues affecting many users, namely: K-Means should be faster for small sample sizes, and the representation of third-party estimators was fixed.<br><br>You can check this version out using:<br><br>    <span style="font-family:monospace">pip install -U scikit-learn</span><br></div><div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">You can see the changelog here: <a href="https://scikit-learn.org/stable/whats_new/v0.23.html#version-0-23-1">https://scikit-learn.org/stable/whats_new/v0.23.html#version-0-23-1</a></div><div><br></div><div><div style="font-size:small" class="gmail_default">The conda-forge builds will be available shortly, which you can then install using:</div><div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default"><span style="font-family:monospace">    conda install -c conda-forge scikit-learn</span></div></div><div><br></div><div><div style="font-size:small" class="gmail_default">On behalf of the scikit-learn development community,</div><div style="font-size:small" class="gmail_default">Adrin<br></div><br></div></div>