Choose a complete project by the type of Machine Learning problem you want to solve. Each example page explains the problem, dataset, important development steps, and relevant Deep Netts API before linking to the runnable Maven project on GitHub.
If you have not run Deep Netts yet, complete Hello World first. Hello World verifies the installation; the projects below demonstrate complete model-development workflows.
If this is your first complete Machine Learning project, start with Simple Linear Regression. It uses one input and one output while still demonstrating data preparation, training, evaluation, model saving, and prediction.
When you are ready to run the projects locally, open the complete guide for cloning the repository, choosing a project, compiling it, enabling the Vector API, and troubleshooting common setup problems.
Each example is a standalone Maven project. Open its directory in any Maven-compatible Java IDE or run it from a terminal using Java 25 with the Vector API module enabled.
For examples that expose the shared training property, pass -Ddeepnetts.epochs=1 as a JVM system property in the IDE run configuration or Maven command.
Small datasets are included under each project's src/main/resources/datasets directory. Large VGG weights are downloaded and cached separately.
Training examples create a local models directory and save the resulting .dnet model there.
