Open nimrodraifer opened 8 years ago
Hi, Just to make sure, is ranking.valid-ndcg-truncation actually reflects the 'k' of NDCG@k used to optimize on the training data ?
Thanks
My reading of the code seems to confirm this... where k <= 10 in jforests
Hi, Just to make sure, is ranking.valid-ndcg-truncation actually reflects the 'k' of NDCG@k used to optimize on the training data ?
Thanks