This practical guide provides more than 200 self-contained recipes to help you solve machine learning challenges you may encounter in your work. If you're comfortable with Python and its libraries, including pandas and scikit-learn, you'll be able to address specific problems, from loading data to training models and leveraging neural networks.

Each recipe in this updated edition includes code that you can copy, paste, and run with a toy dataset to ensure that it works. From there, you can adapt these recipes according to your use case or application. Recipes include a discussion that explains the solution and provides meaningful context.

Go beyond theory and concepts by learning the nuts and bolts you need to construct working machine learning applications. You'll find recipes for:

• Vectors, matrices, and arrays
• Working with data from CSV, JSON, SQL, databases, cloud storage, and other sources
• Handling numerical and categorical data, text, images, and dates and times
• Dimensionality reduction using feature extraction or feature selection
• Model evaluation and selection
• Linear and logical regression, trees and forests, and k-nearest neighbors
• Supporting vector machines (SVM), naäve Bayes, clustering, and tree-based models
• Saving, loading, and serving trained models from multiple frameworks

When the first edition of this book was published in 2018, it filled a critical gap in the growing wealth of machine learning (ML) content. By providing well-tested, hands-on Python recipes, it enabled practitioners to copy and paste code before easily adapting it to their use cases. In a short five years, the ML space has continued to explode with advances in deep learning (DL) and the associated DL Python frameworks.

Chapter 1. Working with Vect ors, Matrices, and Arrays in NumPy

Chapter 2. Loading Data

Chapter 3. Data Wrangling

Chapter 4. Handling Numerical Data

Chapter 5. Handling Categorical Data

Chapter 6. Handling Text

Chapter 7. Handling Dates and Times

Chapter 8. Handling Images

Chapter 9. Dimensionality Reduct ion Using Feature Extract ion

Chapter 10. Dimensionality Reduct ion Using Feature Select ion

Chapter 11. Model Evaluation

Chapter 12. Model Select ion

Chapter 13. linear Regression

Chapter 14. Trees and Forests

Chapter 15. K-Nearest Neighbors

Chapter 16. l ogistic Regression

Chapter 17. Support Vect or Machines

Chapter 18. Naive Bayes

Chapter 19. Clustering

Chapter 20. Tensors with PyTorch

Chapter 21. Neural Networks

Chapter 22. Neural Networks for Unstructured Data

Chapter 23. Saving, Loading, and Serving Trained Models

Now, in 2023, there is a need for the same sort of hands-on content that serves the needs of both ML and DL practitioners with the latest Python libraries. This book intends to build on the existing (and fantastic) work done by the author of the first edition by:

• Updating existing examples to use the latest Python versions and frameworks
• Incorporating modern practices in data sources, data analysis, ML, and DL
• Expanding the DL content to include tensors, neural networks, and DL for text and vision in PyTorch
• Taking our models one step further by serving them in an API

Like the first edition, this book takes a task-based approach to machine learning, boasting over 200 self-contained solutions (copy, paste, and run) for the most common tasks a data scientist or machine learning engineer building a model will run into.

About the Author

Kyle Gallatin is a software engineer for machine learning infrastructure with years of experience as a data analyst, data scientist and machine learning engineer. He is also a professional data science mentor, volunteer computer science teacher and frequently publishes articles at the intersection of software engineering and machine learning. Currently, Kyle is a software engineer on the machine learning platform team at Etsy.

Chris Albon is the Director of Machine Learning at the Wikimedia Foundation, the non-profit that hosts Wikipedia.