Open Zhanelya opened 11 years ago
The issue here seems to be a problem in the implicit model (data training, sklearn). We need a way to closer observe what is happening in that implicit model, because it involves several complicated steps and this is confusing. Because of this we have started the simulation in Orange. One possible outcome could be that once we have a working mechanism in Orange, we can replace the sklearn module with one from Orange.
Even if any incident occur, that is only concerning the fingerprints, which may occur even if no fingerprints are actually used in the policy.. Any idea why that can be?