Developer guides are focused deep-dives into specific Deep Netts development tasks, such as preparing data, configuring training, evaluating a model, or saving it for later use. They are one of the fastest ways to understand how an individual API operation fits into a real Java Machine Learning workflow.
Each guide contains a compact Java recipe, explains the important choices, and links to complete runnable examples. The examples use standalone Maven projects, so you can copy the relevant snippet into your own workflow or open the full project when you need more context.
For the wider conceptual context, continue with Deep Netts API and Model Development. To see every step assembled into one application, choose a project from Examples.
JSR 381 defines standard Java Machine Learning contracts such as DataSet. Deep Netts provides concrete neural-network functionality and interoperates with those shared types.
Split the dataset first, fit preprocessing only on the training set, and apply the fitted transformation to the test set and future inputs.
Yes. A prediction must use the same transformation learned during training, so attach the fitted normalizer or scaler before saving the model.
Use the Developer Guide for one focused API task. Use Examples when you need to see an entire workflow assembled into a runnable project.
