I might have not fully understood the inner workings of the classifier, but wouldn't it be interesting to allow it to learn/calibrate from user input? Basically, is there any way to apply supervised learning to your classifier?
Hi, thanks for your feedback. The classifier behind the scenes has already been trained with a supervised learning algorithm, so it can be customised/adapted to any user input. This is a business case I explore with each customer in order to properly fit the tool to the particularities of their problem (e.g., the specific salient features that represent their data).
