Meetup Boulder Data Science 2016

Scikit-learn provides a consistent framework that makes applying many of the standard machine learning algorithms fairly straightforward. It is certainly one of the reasons that data analysis is becoming so popular in Python. This tutorial will give an overview of of the types of problems that can be addressed with Scikit-Learn. We will use several code examples to explain some of the basic concepts of machine learning, including classification, clustering, and dimensionality reduction.

Notebooks

Introduction

Supervised Learning

• Regression
• Classification
• Examples

Unsupervised Learning

• Dimensionality reduction with PCA
• Clustering
• Examples

Model Selection

• Cross validation
• Bias-Variance Trade-Off

Regularization

• Ridge
• Lasso

Conclusion

Setup

I would install the latest python package from Anaconda. This should have everything you need to run the notebook examples. Here are the versions I am using:

sklearn

>>> import sklearn
>>> sklearn.__version__
'0.16.1'

numpy

>>> import numpy
>>> numpy.__version__
'1.10.4'

matplotlib

>>> import matplotlib
>>> matplotlib.__version__
'1.4.3'

pandas

>>> import pandas
>>> pandas.__version__
'0.17.1'