sigmoid.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A social space for people researching, working with, or just interested in AI!

Server stats:

608
active users

#Sklearn

0 posts0 participants0 posts today

I just did my first project using the #mlflow library to track metrics on iterations of manual tuning of an #sklearn pipeline, it works great and gives me some idea of the search space before moving into automated hyperparameter tuning.

I am using it in a super basic way, as an alternative to creating a gazillion cells with comments tracking metrics, does anyone have any favorite features to check out for taking mlflow to the next level?
#machinelearning #python #MLOps #scikitlearn

In my job as a data analyst, I come across many different types of problems to solve. Some are relatively easy to solve, others not so much. That was until recently, where I came across a problem I have never given much thought before. That was until now.

What is the problem? Finding multiple peaks in a dataset.

You might think, this sounds […]

https://jrashford.com/2024/03/25/finding-peaks-in-a-dataset-and-why-it-is-not-straightforward/

Discover scikit-learn 1.4 and its:
🟢 5 major features & 13 features
🔵 14 efficiency improvements & 23 enhancements
🟡 15 API changes
🔴 38 fixes

More details in the changelog: bit.ly/3tWlZA3
or in the release highlights: bit.ly/3Hsoddm

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 again to all the +80 contributors!

I ran a quick Gradient Boosted Trees vs Neural Nets check using scikit-learn's dev branch which makes it more convenient to work with tabular datasets with mixed numerical and categorical features data (e.g. the Adult Census dataset).

Let's start with the GBRT model. It's now possible to reproduce the SOTA number of this dataset in a few lines of code 2 s (CV included) on my laptop.

1/n

🧵 I recently dived into a rabbit hole when attempting to fix the tests for 's OLS and Ridge regression solvers.

On the theoretical side, I now understand that the minimum norm solution for the centered problem without intercept is also the minimum norm solution for the original problem (with intercept). Ridge/OLS on centered X & y followed by intercept computation is the approach (hereafter name type "a") we have been using for years.

raw.githubusercontent.com/ogri