What is the advantage of utilizing the meta-classifier? When did we decide to use or not to use?

[Evelina CAMILLERI]

 [webinar - Program decision trees in sensors with a Machine Learning Core]

[Evelina CAMILLERI]

The meta-classifier is a filter on the decision tree results. It can be used to fine tune the classification through the usage of dedicated counters.

By setting these counters the number of transitions on the MLC outputs, and consequently the false positives, can be reduced.