Project WP #654: WP6 - Supporting Blue Economy: VREs Development [Months: 1-30]
Project Task #655: T6.1 Performance evaluation, benchmarking and decision making in aquaculture VRE [Months: 1-30]
Project Activity #1648: Blue Economy VRE Cluster #1 Software Development (Stage 2)
|Status:||Closed||Start date:||Sep 22, 2017|
|Assignee:||Denis Pyriochos||% Done:|
The global model get similar sites from the production VREs and then get related models in order to calculate the global KPIs.
AquacultureTrainingLab VRE should have special treatment regarding site similarity. It should be able to get similar sites from the other VREs (the actual Farms) but its sites should not be included in any analysis as the data are not "production" but "training" eg "dummy".
A simple exclusion from the final set of the KPIs that contribute to the global model is not adequate. We should exclude the training sites as soon as possible in the process because of the various grades we attribute to the similarity. When one asks for a global model, the first request is for similar sites asking for "at least X sites of grade Z" (ie "at least 2 sites of grade A", or "at least 5 sites of grade C". If we take into account the training sites, the global model is tricked to think it has more "similar" sites than it actually does.