https://github.com/fokoa/Fault-Impact-Analysis-Towards-Service-Oriented-Network-Operation-Maintenanceby-ITU
Did you combine the KPIs for the train nodes into one node? I can see your train set having 800K+ rows. Can you explain why you did this?
I did this to train the models on all observations of nodes. But I did all my analysis and feature extractions based on nodes.
Did you train the model separately on each node!? kind you trained model 7256 (number of nodes in train) times
Did you combine the KPIs for the train nodes into one node? I can see your train set having 800K+ rows. Can you explain why you did this?
I did this to train the models on all observations of nodes. But I did all my analysis and feature extractions based on nodes.
Did you train the model separately on each node!? kind you trained model 7256 (number of nodes in train) times