Table Of Contents


Built with 🛠 MkDocs - Theme 🖤 Github.

Getting Started

This page takes you from a Java development environment to a verified Deep Netts installation, then points you to the right learning material. Follow the first four sections in order.

1. Prerequisites

The course is written for Java developers who are new to Machine Learning. You do not need previous ML experience, but you should be comfortable reading Java classes, methods, arrays, and Maven project files.

Install or configure:

1.1. Verify Java 25

All runnable Maven projects compile with Java 25:

<maven.compiler.release>25</maven.compiler.release>

Confirm the active runtime and compiler:

java --version
javac --version

Both commands should report Java 25.

1.2. Verify Maven

Each example is a standalone Maven project with its own pom.xml. Confirm Maven is available:

mvn --version

The output includes the Java version and Java home used by Maven. If several JDKs are installed, make sure Maven also reports Java 25.

1.3. Understand the Vector API requirement

Deep Netts 4 uses the incubating Java Vector API. It must be enabled when running an application:

--add-modules jdk.incubator.vector

The Installation section shows how to apply this option from a terminal, an IDE, or a build-tool run configuration.

Add the Vector API option to the VM options (or JVM arguments) of the run configuration that starts the application. The exact field name depends on the IDE or build tool, but the option must be passed to the Java runtime.

2. Download

To run Deep Netts applications, download the distribution used by the examples. You can read this course online without cloning its source repository.

2.1. Choose the compatible distribution

The runnable projects currently require these Maven artifacts:

com.deepnetts:deepnetts-core-pro:4.0.0
com.deepnetts:deepnetts-license:1.0

Download Deep Netts through the official Deep Netts download page. Use the distribution and license provided for your Deep Netts account or organization.

Deep Netts Community Edition is available separately from Maven Central, but it uses the deepnetts-core artifact. The course projects use deepnetts-core-pro; do not substitute one artifact for the other and assume the code will work unchanged.

Before continuing, make sure you have:

Keep the distribution, license, credentials, and proprietary libraries outside source control.

3. Installation

These steps configure Deep Netts for Maven-based Java applications. Installing the optional Visual AI Builder is not required for the course examples.

3.1. Extract the distribution

Extract the downloaded package to a secure local directory. Keep the distribution and license files outside your application repositories.

3.2. Install the library in your local Maven repository

The distribution provides a script that imports the library and license artifacts into your local Maven repository.

On Windows:

ImportToLocalMaven.bat

On Linux or macOS:

chmod +x importToLocalMaven.sh
./importToLocalMaven.sh

Run the script from the extracted distribution directory. The artifacts are normally installed under:

Windows: C:\Users\YOUR_USER\.m2\repository\com\deepnetts
Linux/macOS: ~/.m2/repository/com/deepnetts

Confirm that these versions exist:

com/deepnetts/deepnetts-core-pro/4.0.0/
com/deepnetts/deepnetts-license/1.0/

3.3. Add the Maven dependencies

<dependencies>
    <dependency>
        <groupId>com.deepnetts</groupId>
        <artifactId>deepnetts-core-pro</artifactId>
        <version>4.0.0</version>
    </dependency>

    <dependency>
        <groupId>com.deepnetts</groupId>
        <artifactId>deepnetts-license</artifactId>
        <version>1.0</version>
    </dependency>
</dependencies>

Use the artifact versions supplied with your distribution. To run the examples unchanged, they must match the versions above.

3.4. Target Java 25

<properties>
    <maven.compiler.release>25</maven.compiler.release>
</properties>

Run mvn --version and confirm that Maven uses Java 25.

3.5. Enable the Vector API

On Linux or macOS:

MAVEN_OPTS="--add-modules jdk.incubator.vector" mvn exec:java

On Windows Command Prompt:

set MAVEN_OPTS=--add-modules jdk.incubator.vector
mvn exec:java

On Windows PowerShell:

$env:MAVEN_OPTS="--add-modules jdk.incubator.vector"
mvn exec:java

In an IDE, add this VM option to the application's run configuration:

--add-modules jdk.incubator.vector

3.6. Installation troubleshooting

If Maven cannot resolve an artifact, check that the import script succeeded, the required versions exist under .m2/repository/com/deepnetts, and pom.xml uses matching versions.

If the Vector API module is missing, confirm that Maven runs on Java 25 and the --add-modules jdk.incubator.vector option is applied to the process that starts the application.

4. Hello World

Use the minimal Hello World project to verify the Java and Deep Netts runtime before starting a complete Machine Learning example. It loads Deep Netts 4 Pro and its license, creates a neural network, and runs a real prediction rather than merely printing a greeting.

4.1. What this verifies

A successful run confirms that:

4.2. Get the project

Clone the repository and enter the standalone Hello World project:

git clone https://github.com/deepnetts/cookbook-examples.git
cd cookbook-examples/examples/hello-world

The course website remains available online. Cloning is needed here only because you are about to compile and run a local Java project.

4.3. Confirm the active JDK

mvn --version

The output must report Java 25. If it reports another version, select the correct JDK before continuing.

4.4. Compile

mvn compile

This verifies that Maven can resolve the locally installed Deep Netts Pro and license artifacts and compile the Java source.

4.5. Run

On Linux or macOS:

MAVEN_OPTS="--add-modules jdk.incubator.vector" mvn clean compile exec:java

On Windows Command Prompt:

set MAVEN_OPTS=--add-modules jdk.incubator.vector
mvn clean compile exec:java

On Windows PowerShell:

$env:MAVEN_OPTS="--add-modules jdk.incubator.vector"
mvn clean compile exec:java

4.6. Expected result

The program prints a numerical prediction followed by output similar to:

Java version: 25...
Vector API enabled: true
Deep Netts prediction for 1.0: ...
SUCCESS: Deep Netts is installed and configured correctly.

The exact Java patch version and prediction are not important. Vector API enabled: true confirms the runtime module. Reaching the success message confirms that the library and license loaded, the network was created, and inference completed.

4.7. If Hello World fails

5. Learning resources

Are you a Java developer looking for a concise introduction to Machine Learning with Deep Netts? Start with Machine Learning Basics to learn how samples, features, targets, regression, and classification connect to Java code.

Want to understand Deep Netts 4 and its main capabilities? Read the Deep Netts API lesson for an overview of datasets, preprocessing, model creation, training, evaluation, saving, and prediction.

Are you looking for focused guides covering in-depth usage of individual parts of the Deep Netts API? Open the Deep Netts developer guides and choose the task you want to add to your workflow.

Are you looking for complete tutorials showing Deep Netts in action across different Machine Learning problems? Browse the Deep Netts examples for explained regression and classification walkthroughs linked to runnable Java 25 Maven projects.

The cookbook projects target Java 25 and use Maven. Confirm that both java -version and mvn -version report the expected JDK before running an example.

Deep Netts uses the JDK Vector API for optimized numerical operations. Pass --add-modules jdk.incubator.vector to the Java process as a runtime argument.

Yes. Each example is a standard Maven project. Use Java 25, pass the Vector API option to the Java process through your IDE's run configuration or the command line, and run the main class.

Use the Free Download link in the header, then follow the Download and Installation sections on this page to add the required artifacts and license.