ML in Java, YES it's possible! by Mohammed Aboullaite

Discover how to leverage Java for machine learning, including deploying models on edge devices, and explore popular libraries like DGL and Deep Learning 4G.

Key takeaways

Java can be used for machine learning, and it’s possible to deploy it on edge devices.
Artificial intelligence needs a lot of data to learn patterns.
Traditional programming involves writing rules to get the desired output, while machine learning involves feeding the model with data and letting it learn.
Labels are the output of the model, and they are used to train it.
Machine learning engineers spend a lot of time tweaking model parameters to get the best results.
Using Java for machine learning allows for more interoperability with other platforms.
There are many machine learning libraries available for Java, including DGL (Deep Graph Library) and Deep Learning 4G.
LibIndie4G is a powerful engine for building machine learning models.
The model zoo is a collection of pre-trained models that can be used directly.
Deep learning involves using multiple layers to build complex models.
Generative AI focuses on generating new output, not just predicting existing data.
Java has a strong community and many resources available for machine learning.
Deep Learning 4G is a library for building machine learning models in Java.
The reference implementation for DGL is available as part of the OpenJSR project.
Oracle Tribute is a machine learning platform that supports multiple data sources and allows for easy deployment.
Java is used in many industries, including finance, credit card processing, and security.
Java has been used to build many successful applications, including those for image recognition, natural language processing, and fraud detection.
The key feature of Deep Learning 4G is its support for multiple deep learning frameworks.
The model serving part of Oracle Tribute allows for easy deployment of machine learning models.
Generative AI is a new field that is focused on generating new output, not just predicting existing data.
The author’s favorite machine learning library is DGL, which is easy to use and has a strong model zoo.
The presentation includes many examples of how to use Java for machine learning, including image recognition and natural language processing.
The key to successfully using machine learning is to provide the model with enough data to learn patterns.

ML in Java, YES it's possible! by Mohammed Aboullaite

More talks