A cookbook for TensorFlow’s graph-and-placeholder era

What it does

This repo is the code companion for Nick McClure’s Packt book, TensorFlow Machine Learning Cookbook. It houses eleven chapters of examples—from basic tensors and linear regression through CNNs, RNNs, and NLP—implemented with explicit TensorFlow operations. The README serves as a visual table of contents, linking each recipe to its own folder and thumbnail.

The interesting bit

The value is in the verbosity: every recipe builds explicit computational graphs, declares placeholders, and wires back-propagation by hand. It is less a reusable library and more a dissected lab notebook for understanding how the pieces fit together under the hood.

Key highlights

• Covers a full ML curriculum: linear and logistic regression, SVMs, nearest neighbors, neural networks, CNNs, RNNs, bag-of-words, word2vec, genetic algorithms, and ODEs.
• Mirrors the book structure exactly, with per-chapter folders and inline thumbnails for each recipe.
• Includes chapters on production deployment and auxiliary topics like TensorBoard, k-means clustering, and genetic algorithms.
• Uses TensorBoard throughout to visualize graph structures, loss curves, and layer layouts.

Caveats

• The README is almost entirely a table of contents; actual implementation details live inside the chapter folders and are not excerpted.
• The code leans heavily on placeholders and manual computational graphs, so it is verbose by design and not a high-level API wrapper.

Verdict

Worth bookmarking if you are working through the Packt book or want to see machine learning algorithms constructed from explicit TensorFlow operations. Less useful if you need a standalone library to import.