Train Models In-Memory On Your Operational Dataset
GridGain® machine learning is part of the GridGain Continuous Learning Framework which enables model training in real-time on your operational dataset which can improve models and outcomes as events happen. It accomplishes this by including machine learning libraries directly in GridGain. These libraries are optimized for massively parallel processing (MPP), which allows the system to run each algorithm locally against the data residing in memory on each node of the GridGain cluster.
Reduced Machine Learning Time and Cost
Processing your operational data in place in GridGain eliminates the time and cost required to move transactional data across a network into a separate machine learning database before model training. It also takes advantage of GridGain’s in-memory speed and unlimited horizontal scalability. The result is real-time model training at petabyte scale with a much lower cost than traditional approaches.
Continuous Machine Learning
With GridGain machine learning, models can be continuously retrained as events happen to help improve decisions and outcomes. GridGain supports hybrid transactional/analytical processing (HTAP) by allowing any analytics or machine learning to be run in-place in memory within a transaction or interaction. It is integrated with a host of streaming technologies – including Apache® Camel™, Apache® Flink™, Apache® Flume™, Apache® Kafka™, Apache® RocketMQ™, Apache® Spark™, Apache® Storm™, Java Message Service (JMS), MQTT, Twitter and ZeroMQ – to ingest and process data streams in real-time. These libraries can be accessed directly by any developer using GridGain’s unified API that supports many different languages.