Say I have some process setup and it's working like I want it to on my test data. Now I want to classify new data. But not only that, I would like to do it in real-time within RapidMiner. Is this possible to do? I envision that whenever an example set is updated with new examples, it would run the algos only on the new data and update the output resultSet accordingly.
using RapidMiner alone this is probably not possible. The reason is not that you cannot update your predictions which is certainly possible. However, I don't see how you would trigger the event of new examples arriving. Where are they coming from? I think the best way would be to design some service-like thing, e.g. deploying your process in RapidAnalytics and classifying examples using the service.
Best,
Simon