While running the clustering solution, by default 50% of the records are considered for training the solution.
I tried using the Advanced Solution setting : Target Solution Coverage = 60%/80%/100%.
The records which were excluded from training the clustering solution when the threshold is 50%, were also part of the missing records when using the Target Solution Coverage = 60%/80%.
By this we can understand that the records are not excluded from training randomly.
Can anybody help me understand what logic is being used for exclusion of records?
Thankyou,
Chaitra
