Hi Team,
I didn’t understand the sentence ‘In this particular case, only the first 2 decimals seem to be trustworthy.’ which was mentioned in the notebook.
Next question is:
The mean cross-validation accuracy is: 0.800 +/- 0.003
what to do if the second decimal value goes high? let’s say +/- 5.452
It means that you results will vary between 0.800 - 3 * 0.003 = 0.791 and 0.800 + 3 * 0.003 = 0.809.
So providing results after the 3rd digits will be useless.
If you have a large variation, it means that you don’t even need decimal anymore.
So, if we have a large variation. It means the model isn’t consistent and it wont generalise well on the new data. Correct me if i’m wrong.
Yes kinda. It is even worse, it means that you have no idea what it will do on new unseen data 
Yeah, right. Thanks very much