Train Deep Learning Models Using Operational Datasets

The GridGain® deep learning capability is part of the GridGain Continuous Learning Framework which enables real-time model training and can improve models and outcomes as events happen. The deep learning capability is built on the Apache Ignite multilayer perceptron neural network which is included in GridGain. The neural network is optimized for massively parallel processing (MPP), allowing the system to run each algorithm locally against the data residing on each node of the GridGain cluster.

Reduced Deep Learning Time and Cost

Processing the data in place eliminates the time and cost of moving transactional data into a separate deep learning database before model training. The deep learning capability also takes advantage of the in-memory speed and massive horizontal scalability of GridGain. The system offers real-time model training at petabyte scale and does so at a much lower cost than approaches which rely on separate operational and analytical databases.

Continuous Deep Learning

Models can be continuously retrained as events happen to help improve decisions and outcomes with GridGain deep learning. The GridGain in-memory computing platform supports hybrid transactional/analytical processing (HTAP), allowing any analytics or deep learning to be run in-place in memory within a transaction or interaction. GridGain is integrated with many common 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 which allow the system to ingest and process data streams in real-time. These libraries can be accessed directly by any developer using the GridGain Unified API that supports many different languages.