#playground-series-s3e22
1 messages Β· Page 1 of 1 (latest)
Well, at least it's good to know someone's got a lot of free time π
Yes I also get a downvote for all my discussions.π
All of us get downvoted at some point or not. Like it or not, it is a feedback about what we are doing. I think it is a good idea to heed that feedback, except on those cases when it is a single downvote without any clue as to why it happened.
Itβs a mass downvote, even those with gold medals as well
Your discussion about the "mass downvoter" also got downvoted
That person is clearly not respecting your discussion post that exposed him
Hi @stable oriole , thanks for bringing this to our attention! Wanted to let you know we are aware of this issue and will take appropriate action.
hello everyone
please, how to import our dataset in google collab?
you can download it and upload to google drive and then load your drive in the notebook to make your life easier when you close and open the notebook ,
try to search the web or youtube if you face minor issues like this , that way you get faster responses and improve your researching capabilities
ok thank
Hey Peeps,
I am a begineer and trying to find a Datadict for the description of the columns. It would be amazing if someone can provide me with the link to the datadict.
You can use kaggle static API for this
No need to download and upload
1 public notebook is there of ps3e22, having public LB as .87915, but 1 submission file is missing there, & while running, it is giving public LB as .81707 instead of .87915. Has anyone noticed that?
Take note of the version / revision. Their best score may have been a previous attempt
I have tried all the versions, but none of them giving public LB score as 0.87195.
Will this sustain??
I mean I think we will have a churn this time
I'm expecting ICR style shakeup on this one
As private LB is 80 percent of the data, so I think churn will be definite.
In 5.5 hours we will know if this score sustains
In my opinion, the winner is lurking in positions 450-550
I fear for the scores above 0.83
Yeah
It seems like across the board models find it hard to predict some false positives for live live and die die
@uneven talon do you know if there has been shakeups in past playground comps
How should i evaluate my test data score like actually from train have less samples and from test predic have more samples
