Sold all my crypto and used that in part to buy a house. I don’t have the appetite to take more risks in crypto. But of all the cryptocurrencies I am most optimistic about NMR and my models seem to be performing well.
3 Likes
… Isn’t the example model included in dataset (the one using xgboost) perform better than your commercialized model? sorry if I am wrong, I don’t know much details but just look at the corr and mmc value…
1 Like
Hi gockgang! The live results are close enough, but the number of finished rounds for my model is too low for any strong conclusions (jackerparker4 is live only from round 263). The local CV and validation stats are much better for my model compared to example predictions. But in general your question is good and I don’t have enough live data to prove that my model is much better what I’m trying to compensate by collateral information and low price right now.
Regards,
Mark
1 Like