Dive into the world of hands-on machine learning with this curated collection of notebooks, designed to guide you through the fundamentals of Machine Learning in Python. These notebooks serve as practical companions to the acclaimed book, “Hands-on Machine Learning with Scikit-Learn, Keras, and TensorFlow,” offering you example code and solutions to exercises directly from the second edition.
For those seeking the most up-to-date resources, the notebooks for the 3rd edition are available at ageron/handson-ml3, featuring even more current code and practices. If you are interested in the notebooks from the first edition, you can find them at ageron/handson-ml.
Getting Started with Hands-On Machine Learning
Embarking on your hands-on machine learning journey is straightforward. Choose the method that best suits your learning style and technical setup:
Learn Interactively Online
Want to start coding immediately without any installation hassles? Leverage these online platforms to run the notebooks directly in your browser:
- Google Colab: Offers free GPU and TPU resources, ideal for computationally intensive tasks.
- Kaggle Kernels: Another excellent platform providing free computing power and a collaborative environment.
Important Note: Remember that these cloud environments are temporary. Be sure to download any work or data you wish to save.
Quick Notebook Preview
If you’re just looking to browse the code and explore the notebooks without executing them, GitHub’s notebook viewer provides a convenient way to read through the content directly online.
Run Locally with Docker
For a consistent and pre-configured environment, consider using the Docker image for this project. Detailed instructions are available in the Docker instructions directory.
Local Installation for In-Depth Exploration
To fully immerse yourself in hands-on machine learning and run the notebooks on your own machine, follow these steps:
-
Install Prerequisites: Begin by installing Anaconda or Miniconda, git. If you have a compatible GPU for TensorFlow, install the GPU driver along with the appropriate CUDA and cuDNN versions, as detailed in TensorFlow’s documentation.
-
Clone the Repository: Open your terminal and execute these commands to clone the project:
$ git clone https://github.com/ageron/handson-ml2.git $ cd handson-ml2 -
Environment Setup: Run these commands to create and activate the necessary conda environment:
$ conda env create -f environment.yml $ conda activate tf2 $ python -m ipykernel install --user --name=python3 -
Launch Jupyter: Finally, start Jupyter Notebook to access the notebooks:
$ jupyter notebook
For more detailed guidance, refer to the comprehensive installation instructions.
Frequently Asked Questions
Which Python version is recommended?
Python 3.8 is the recommended version for optimal compatibility. Following the installation steps will set up this version for you. While most code should function on other Python 3 versions, Python 3.8 ensures the broadest library support.
Troubleshooting load_housing_data() errors:
Ensure you have executed fetch_housing_data() before calling load_housing_data(). For HTTP errors, meticulously compare your code to the notebook examples, copying and pasting if necessary. Network configuration issues may also be a cause.
Resolving SSL errors on MacOSX:
SSL certificate issues on MacOSX often require manual installation. Consult this StackOverflow question for solutions. If using the official Python distribution, run /Applications/Python 3.8/Install Certificates.command (adjust version number as needed). MacPorts users can try sudo port install curl-ca-bundle.
Updating your local project to the latest version:
Refer to the INSTALL.md file for instructions on updating your local installation.
Updating Python libraries with Anaconda:
Guidance on updating Python libraries within your Anaconda environment can also be found in the INSTALL.md file.
Contributing to Hands-On Machine Learning Education
This project thrives on community contributions. We extend our gratitude to all contributors who have provided valuable feedback, reported issues, and submitted pull requests. Special thanks to Haesun Park and Ian Beauregard for their thorough notebook reviews and PR contributions, including exercise solution assistance. Additionally, thank you to Steven Bunkley and Ziembla for creating the docker directory, and to GitHub user SuperYorio for contributions to exercise solutions.
By engaging with these hands-on machine learning notebooks, you are taking a significant step towards mastering practical machine learning skills. Start exploring today and transform your theoretical knowledge into real-world expertise.
