Book Image

Python Data Science Essentials - Second Edition

By : Luca Massaron, Alberto Boschetti
Book Image

Python Data Science Essentials - Second Edition

By: Luca Massaron, Alberto Boschetti

Overview of this book

Fully expanded and upgraded, the second edition of Python Data Science Essentials takes you through all you need to know to suceed in data science using Python. Get modern insight into the core of Python data, including the latest versions of Jupyter notebooks, NumPy, pandas and scikit-learn. Look beyond the fundamentals with beautiful data visualizations with Seaborn and ggplot, web development with Bottle, and even the new frontiers of deep learning with Theano and TensorFlow. Dive into building your essential Python 3.5 data science toolbox, using a single-source approach that will allow to to work with Python 2.7 as well. Get to grips fast with data munging and preprocessing, and all the techniques you need to load, analyse, and process your data. Finally, get a complete overview of principal machine learning algorithms, graph analysis techniques, and all the visualization and deployment instruments that make it easier to present your results to an audience of both data science experts and business users.
Table of Contents (13 chapters)
Python Data Science Essentials - Second Edition
Credits
About the Authors
About the Reviewer
www.PacktPub.com
Preface

An overview of unsupervised learning


In all the methods we've seen so far, every sample or observation has its own target label or value. In some other cases, the dataset is unlabeled and, in order to extract the structure of the data, you need an unsupervised approach. In this section, we're going to introduce two methods to perform clustering, as they are among the most used methods for unsupervised learning.

Note

It is useful to keep in mind that often the terms "clustering" and "unsupervised learning" are considered synonymous, though actually unsupervised learning has a larger meaning.

The first method that we'll introduce, named K-means, is the most commonly used clustering algorithm despite its inevitable shortcomings. In signal processing, K-means is the equivalent of a vectorial quantization, that is, the selection of the best code word (from a given codebook) that better approximates the input observation (or a word).

You must provide the algorithm with the K parameter, which is the...