Home List Contact
Welcome to the machine learning kernels repository!

Machine learning is the study of teaching machines to learn using data, with and without human supervision.

Some time ago I (@residentmario) decided that I needed to systematize my knowledge of the field by studying it from the ground up. I decided to use kernels to do it.

Kernels are a wonderful utility from the folks @Kaggle that provide an editable notebook environment that anyone else can then freely fork and run themselves. If you're familiar with notebooks, they're just like that, except on the cloud.

This microsite collects the write-ups I did into an organized, readable list. I cover material that ranges from the basic to the intermediate, starting with basic linear regression and train-test splits and ending with topics such as random forests, oversampling, and probability calibration. There are currently fifty-five notebooks in the list!

Implementations are in Python, using the scikit-learn library and friends.

Ready to get started? Check out the list →

Questions, concerns, or errata? Contact me →