Batch Learning
lastmod 2021-02-20created 2021-02-191 edits
In batch learning, the system cannot learn incrementally. It must be trained using all of the available data.
Because this approach consumes a lot of time and resources, it is usually run offline. The system is trained first and, once deployed to the production system, runs without any further learning. In other words, it only applies what it has learned. This is called offline learning (Offline Learning).
For a batch learning system to learn from new data, a new version of the system must be trained from scratch on the full data. The old system is then stopped and replaced with the new one.
This approach is simple and works well, but training on the full dataset can take several hours. Also, because it trains on the full dataset, it consumes a lot of system resources. When a resource-constrained system has to learn on its own, hauling large amounts of training data and spending resources on training can cause problems.