sklearnsk 0.1.0

A wrapper for sklearn, that makes it easier to write, tune and evaluate classification and regression systems

A sidekick for scikit-learn that makes it easier to write, tune and evaluate classification and regression systems

Installation

Install from the python package index:

pip install sklearnsk

Or clone this repository and install:

pip install .

Usage

Check out the following notebooks in the example directory for examples of usage:

• iris.ipynb: A toy classification problem

• boston.ipynb: A toy regression problem

• 20newsgroups.ipynb: A more complex classification problem, involving n-grams, one-hot encoding, feature selection, etc.

Each of these examples will take you through the process of defining your system, tuning it (with some nice visualisation), evaluating it, and performing additional analysis like feature ablation.

Licence

This code is released under the MIT licence