![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F347463%2F43f5a70a35e773d4153858f818019e82%2Friiid-test-answer-prediction-shakeup.png?generation=1610068750477129&alt=media\")
Guys, it looks we miss one usual thread! Shake up or not shake up for this competition?
\nI think (hope) shake up will be low because:
Guys, it looks we miss one usual thread! Shake up or not shake up for this competition?
\nI think (hope) shake up will be low because:
Fairly small shakeup, I guess. I would add that many people reported their intermediate private score and from what I could see it was in the +- 0.002 range
", "rawMarkdown": "Fairly small shakeup, I guess. I would add that many people reported their intermediate private score and from what I could see it was in the +- 0.002 range", "votes": null }, { "id": "1143198", "postDate": "01/07/2021 19:35:57", "content": "Agreed, my CV / LB been consistent throughout all of my submissions, it is very unlikely that private dataset score is any different.
", "rawMarkdown": "Agreed, my CV / LB been consistent throughout all of my submissions, it is very unlikely that private dataset score is any different.", "votes": null }, { "id": "1143240", "postDate": "01/07/2021 20:02:33", "content": "How soon after deadline is the private leaderboard revealed? First competition for me, this is all new :) Thanks
", "rawMarkdown": "How soon after deadline is the private leaderboard revealed? First competition for me, this is all new :) Thanks", "votes": null }, { "id": "1143245", "postDate": "01/07/2021 20:04:43", "content": "Immediately. Then some cleanup could be done if Kaggle finds some cheaters and usually after a few days it's frozen.
", "rawMarkdown": "Immediately. Then some cleanup could be done if Kaggle finds some cheaters and usually after a few days it's frozen.", "votes": null }, { "id": "1143248", "postDate": "01/07/2021 20:05:58", "content": "@darren1515, private LB is revealed immediately but finalized a few days later after all cheaters are removed and results verified.
", "rawMarkdown": "darren1515, private LB is revealed immediately but finalized a few days later after all cheaters are removed and results verified.", "votes": null }, { "id": "1143256", "postDate": "01/07/2021 20:11:18", "content": "Thanks both
", "rawMarkdown": "Thanks both", "votes": null }, { "id": "1143272", "postDate": "01/07/2021 20:22:10", "content": "@mpware Could I submit before deadline, but leave the notebook running for grading after the deadline? Would it be considered for my final score ?
", "rawMarkdown": "mpware Could I submit before deadline, but leave the notebook running for grading after the deadline? Would it be considered for my final score ?", "votes": null }, { "id": "1143298", "postDate": "01/07/2021 20:39:08", "content": "@abdessalemboukil, your final selections must finish running before the dateline i.e. only choose submissions with public scores showing as your final 2.
", "rawMarkdown": "abdessalemboukil, your final selections must finish running before the dateline i.e. only choose submissions with public scores showing as your final 2.", "votes": null }, { "id": "1143301", "postDate": "01/07/2021 20:41:22", "content": "@abdessalemboukil no, it's a bit late now except if your submission can finish within 3 hours.
", "rawMarkdown": "abdessalemboukil no, it's a bit late now except if your submission can finish within 3 hours.", "votes": null }, { "id": "1143306", "postDate": "01/07/2021 20:45:29", "content": "Oh that sucks! I got my model improved by 0.002 , the specifications of the competition confused me. Thanks for the clarification though!
", "rawMarkdown": "Oh that sucks! I got my model improved by 0.002 , the specifications of the competition confused me. Thanks for the clarification though!", "votes": null }, { "id": "1143527", "postDate": "01/08/2021 00:10:29", "content": "So almost no shake up!
", "rawMarkdown": "So almost no shake up!", "votes": null }, { "id": "1143631", "postDate": "01/08/2021 01:21:46", "content": "Well, I didn't take part, but ran my script anyway out of curiosity: here's the plot, not quite worthy of a new topic…
\n
The top shake-ups ranked by scatter plot file size are here.
", "rawMarkdown": "Well, I didn't take part, but ran my script anyway out of curiosity: here's the plot, not quite worthy of a new topic...\n\n\n\nThe top shake-ups ranked by scatter plot file size are [here][1].\n\n[1]: https://www.kaggle.com/jtrotman/meta-kaggle-scatter-plot-competition-shake-up", "votes": null }, { "id": "1143898", "postDate": "01/08/2021 05:56:45", "content": "yes almost no shake up, is it always no shake up like this if competition has big amount of data?
", "rawMarkdown": "yes almost no shake up, is it always no shake up like this if competition has big amount of data?", "votes": null }, { "id": "1144071", "postDate": "01/08/2021 08:11:01", "content": "No, it depends also of the quality of data (bad labels …) and split public/private.
", "rawMarkdown": "No, it depends also of the quality of data (bad labels ...) and split public/private.", "votes": null } ], "comments": [ { "id": 1143109, "author_name": "stecasasso", "author_url": "", "post_date": "01/07/2021 18:57:54", "content": "Fairly small shakeup, I guess. I would add that many people reported their intermediate private score and from what I could see it was in the +- 0.002 range
", "votes": null, "replies": [ { "id": 1143198, "author_name": "abdessalemboukil", "author_url": "", "post_date": "01/07/2021 19:35:57", "content": "Agreed, my CV / LB been consistent throughout all of my submissions, it is very unlikely that private dataset score is any different.
", "votes": null, "replies": [] } ] }, { "id": 1143240, "author_name": "darren1515", "author_url": "", "post_date": "01/07/2021 20:02:33", "content": "How soon after deadline is the private leaderboard revealed? First competition for me, this is all new :) Thanks
", "votes": null, "replies": [ { "id": 1143245, "author_name": "mpware", "author_url": "", "post_date": "01/07/2021 20:04:43", "content": "Immediately. Then some cleanup could be done if Kaggle finds some cheaters and usually after a few days it's frozen.
", "votes": null, "replies": [] }, { "id": 1143248, "author_name": "sheriytm", "author_url": "", "post_date": "01/07/2021 20:05:58", "content": "@darren1515, private LB is revealed immediately but finalized a few days later after all cheaters are removed and results verified.
", "votes": null, "replies": [] }, { "id": 1143256, "author_name": "darren1515", "author_url": "", "post_date": "01/07/2021 20:11:18", "content": "Thanks both
", "votes": null, "replies": [] }, { "id": 1143272, "author_name": "abdessalemboukil", "author_url": "", "post_date": "01/07/2021 20:22:10", "content": "@mpware Could I submit before deadline, but leave the notebook running for grading after the deadline? Would it be considered for my final score ?
", "votes": null, "replies": [] }, { "id": 1143298, "author_name": "sheriytm", "author_url": "", "post_date": "01/07/2021 20:39:08", "content": "@abdessalemboukil, your final selections must finish running before the dateline i.e. only choose submissions with public scores showing as your final 2.
", "votes": null, "replies": [] }, { "id": 1143301, "author_name": "mpware", "author_url": "", "post_date": "01/07/2021 20:41:22", "content": "@abdessalemboukil no, it's a bit late now except if your submission can finish within 3 hours.
", "votes": null, "replies": [] }, { "id": 1143306, "author_name": "abdessalemboukil", "author_url": "", "post_date": "01/07/2021 20:45:29", "content": "Oh that sucks! I got my model improved by 0.002 , the specifications of the competition confused me. Thanks for the clarification though!
", "votes": null, "replies": [] } ] }, { "id": 1143527, "author_name": "mpware", "author_url": "", "post_date": "01/08/2021 00:10:29", "content": "So almost no shake up!
", "votes": null, "replies": [ { "id": 1143898, "author_name": "superchenhao", "author_url": "", "post_date": "01/08/2021 05:56:45", "content": "yes almost no shake up, is it always no shake up like this if competition has big amount of data?
", "votes": null, "replies": [] }, { "id": 1144071, "author_name": "mpware", "author_url": "", "post_date": "01/08/2021 08:11:01", "content": "No, it depends also of the quality of data (bad labels …) and split public/private.
", "votes": null, "replies": [] } ] }, { "id": 1143631, "author_name": "jtrotman", "author_url": "", "post_date": "01/08/2021 01:21:46", "content": "Well, I didn't take part, but ran my script anyway out of curiosity: here's the plot, not quite worthy of a new topic…
\n
The top shake-ups ranked by scatter plot file size are here.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1143098": "Guys, it looks we miss one usual thread! Shake up or not shake up for this competition?\nI think (hope) shake up will be low because:\n- CV/LB was consistent so far\n- We had a lot of data", "1143109": "Fairly small shakeup, I guess. I would add that many people reported their intermediate private score and from what I could see it was in the +- 0.002 range", "1143198": "Agreed, my CV / LB been consistent throughout all of my submissions, it is very unlikely that private dataset score is any different.", "1143240": "How soon after deadline is the private leaderboard revealed? First competition for me, this is all new :) Thanks", "1143245": "Immediately. Then some cleanup could be done if Kaggle finds some cheaters and usually after a few days it's frozen.", "1143248": "darren1515, private LB is revealed immediately but finalized a few days later after all cheaters are removed and results verified.", "1143256": "Thanks both", "1143272": "mpware Could I submit before deadline, but leave the notebook running for grading after the deadline? Would it be considered for my final score ?", "1143298": "abdessalemboukil, your final selections must finish running before the dateline i.e. only choose submissions with public scores showing as your final 2.", "1143301": "abdessalemboukil no, it's a bit late now except if your submission can finish within 3 hours.", "1143306": "Oh that sucks! I got my model improved by 0.002 , the specifications of the competition confused me. Thanks for the clarification though!", "1143527": "So almost no shake up!", "1143631": "Well, I didn't take part, but ran my script anyway out of curiosity: here's the plot, not quite worthy of a new topic...\n\n\n\nThe top shake-ups ranked by scatter plot file size are [here][1].\n\n[1]: https://www.kaggle.com/jtrotman/meta-kaggle-scatter-plot-competition-shake-up", "1143898": "yes almost no shake up, is it always no shake up like this if competition has big amount of data?", "1144071": "No, it depends also of the quality of data (bad labels ...) and split public/private." }, "source": "meta" }