At the height of machine learning, programmers and scientists analyze big data and deliver predictions to change the world. Machine learning is not only indispensable–it’s more accessible than most realize. With the Scikit-learn library, building effective machine learning models can take very few lines of code without compromising accuracy or power.
This tutorial provides the essential code and comprehension to make meaningful predictions using machine learning. Emphasis is placed not just on writing machine learning code, but on understanding what happens behind the code. With sufficient practice and repetition, as given in this tutorial, anyone can build machine learning models.
In the first module, we frame the problem that machine learning attempts to solve in terms of making predictions from data before building our first model with Linear Regression. In the second module, we use machine learning to classify data using Logistic Regression, comparing and contrasting details of regressors and classifiers. In the third module, we build tree-based models to improve predictions while covering crucial machine learning concepts such as overfitting, splitting data, cross-validation, and adjusting parameters.
The goal here is simple: gain enough proficiency with Scikit-learn to confidently build machine learning models on your own. By the end of this tutorial, you should be able to make meaningful predictions from tabular data (supervised learning) that is primarily numerical. All examples are taken from real-world datasets. Everyone receives complete notebooks and extensive resources to carry the journey forward.
If you are comfortable writing functions in Python and accessing libraries you are sufficiently prepared.