As mentioned in the introduction to the chapter, three approaches to machine learning within Storm will be presented. It is important to understand when to use each approach. This choice starts by understanding what machine learning approach and algorithm you would like to use: online or batch-based. Remember that, for machine learning, online and batch refers to the way in which the model is trained. This distinction does not imply any particular underlying engineering approach to achieve either the batch or online modes. A batch model can be built in a real-time software platform, and conversely, an online model can be built as part of a batch software process. It is important to understand this distinction between the engineering aspect and the machine learning aspect.

If you would like to explore this concept in more depth, please read the following overview of neural networks in which the distinction is explored in more detail: http...