15 Machine Learning: Classification, Regression and Clustering

Objectives

In this chapter you’ll:

• Use scikit-learn with popular datasets to perform machine learning studies.

• Use Seaborn and Matplotlib to visualize and explore data.

• Perform supervised machine learning with k-nearest neighbors classification and linear regression.

• Perform multi-classification with Digits dataset.

• Divide a dataset into training, test and validation sets.

• Tune model hyperparameters with k-fold cross-validation.

• Measure model performance.

• Display a confusion matrix showing classification prediction hits and misses.

• Perform multiple linear regression with the California Housing dataset.

• Perform dimensionality reduction with PCA and t-SNE on the Iris and Digits datasets to prepare them for two-dimensional visualizations.

• Perform unsupervised machine learning with k-means clustering and the Iris dataset.

Outline

1. 15.1 Introduction to Machine Learning

   1. 15.1.1 Scikit-Learn

   2. 15.1.2 Types of Machine Learning

   3. 15.1.3 Datasets Bundled with Scikit-Learn

   4. 15.1.4 Steps in a Typical Data Science Study

2. 15.2 Case Study: Classification with k-Nearest Neighbors and the Digits Dataset, Part 1

   1. 15.2.1 k-Nearest Neighbors Algorithm

   2. 15.2.2 Loading the Dataset

   3. 15.2.3 Visualizing the Data

   4. 15.2.4 Splitting the Data for Training and Testing

   5. 15.2.5 Creating the Model

   6. 15.2.6 Training the Model

   7. 15.2.7 Predicting Digit Classes

3. 15.3 Case Study: Classification with k-Nearest Neighbors and the Digits Dataset, Part 2

   1. 15.3.1 Metrics for Model Accuracy

   2. 15.3.2 K-Fold Cross-Validation

   3. 15.3.3 Running Multiple Models to Find the Best One

   4. 15.3.4 Hyperparameter Tuning

4. 15.4 Case Study: Time Series and Simple Linear Regression

5. 15.5 Case Study: Multiple Linear Regression with the California Housing Dataset

   1. 15.5.1 Loading the Dataset

   2. 15.5.2 Exploring the Data with Pandas

   3. 15.5.3 Visualizing the Features

   4. 15.5.4 Splitting the Data for Training and Testing

   5. 15.5.5 Training the Model

   6. 15.5.6 Testing the Model

   7. 15.5.7 Visualizing the Expected vs. Predicted Prices

   8. 15.5.8 Regression Model Metrics

   9. 15.5.9 Choosing the Best Model

6. 15.6 Case Study: Unsupervised Machine Learning, Part 1—Dimensionality Reduction

7. 15.7 Case Study: Unsupervised Machine Learning, Part 2—k-Means Clustering

   1. 15.7.1 Loading the Iris Dataset

   2. 15.7.2 Exploring the Iris Dataset: Descriptive Statistics with Pandas

   3. 15.7.3 Visualizing the Dataset with a Seaborn pairplot

   4. 15.7.4 Using a KMeans Estimator

   5. 15.7.5 Dimensionality Reduction with Principal Component Analysis

   6. 15.7.6 Choosing the Best Clustering Estimator

8. 15.8 Wrap-Up

1. Exercises