Thank you for sharing this interesting approach! I am wondering how the train phase looks like in practice and if it is really more efficient to create a "custom model" that needs to be trained on a specific dataset. Creating and training a model requires a lot of resources and expertise which most companies probably do not have. Maybe you can share your experience and elaborate on how this phase is realised.
@d.g.567
4 ай бұрын
What was the difference to the double diamond except the fact that the last convergence is not time bound and rather progressive based on the decision on how many iterations should be done to come to a conclusion?
Пікірлер: 2