Thoughtful Deep Learning with Python

Deep Learning with Python and Keras, the thoughtful way. Testing and validating Keras projects.

Original examples are taken from the book Deep Learning with Python by Francois Chollet. These are enhanced with methods that are highly influenced by the book Thoughtful Machine Learning with Python by Matthew Kirk.

Content

Project name and purpose of example:

• IMDB_binary_classification - Unittest all the things (incl. anti-pattern of testing the framework)
• Reuters_multiclass_classification - Plot training history and evaluate metrices of final model
• HousePrices_regression - Imperative project setup and K-fold validation
• IMDB_GloVe_classification - Functional project setup, text tokenizations, plot model informations, use pretrained embeddings
• IMDB_CNN_classification - Callbacks and Tensorboard for monitoring training

For more details, see readme.md in each projects main directory.

Requirements

Run conda env create -f environment.yml to create an environment with all dependencies. Afterwards run conda activate DeepLearning to activate it.

License

MIT License