If someone selects that sub, will it count as cheating?
", "messages": [ { "id": 324776, "postDate": "2018-05-07T23:50:54.737Z", "content": "If someone selects that sub, will it count as cheating?
", "rawMarkdown": "If someone selects that sub, will it count as cheating?", "votes": 8 }, { "id": 325277, "postDate": "2018-05-08T09:17:08.747Z", "content": "I personally would be disappointed if people were disqualified for this. What they did may be unethical, but it is not against the rules. Retroactively changing the rules is not cool in my book.
\n\nChanging the rules after the fact happened in Data Science Bowl 2018 and people were upset (rightfully so). Changing the rules shouldn't be allowed here either. The best course of action is to change things for the better moving forward.
", "rawMarkdown": "I personally would be disappointed if people were disqualified for this. What they did may be unethical, but it is not against the rules. Retroactively changing the rules is not cool in my book.\n\nChanging the rules after the fact happened in [Data Science Bowl 2018](https://www.kaggle.com/c/data-science-bowl-2018/discussion/55228) and people were upset (rightfully so). Changing the rules shouldn't be allowed here either. The best course of action is to change things for the better moving forward.", "votes": 3 }, { "id": 325672, "postDate": "2018-05-08T17:40:54.507Z", "content": "No one will be removed for using Dirk's submission.
", "rawMarkdown": "No one will be removed for using Dirk's submission. ", "votes": 2, "replies": [ { "id": 325689, "postDate": "2018-05-08T18:05:36.800Z", "content": "gg
", "rawMarkdown": "gg", "votes": 1 }, { "id": 326486, "postDate": "2018-05-09T20:04:24.637Z", "content": "@inversion\nBecause of the huge data and long time for training, it is just almost no chance to get big improvement in 5 hours.\nSo I suggest that find the timestamp of Dirk's submission, revert any submissions after that with huge improvement to their previous score.
", "rawMarkdown": "@inversion\nBecause of the huge data and long time for training, it is just almost no chance to get big improvement in 5 hours.\nSo I suggest that find the timestamp of Dirk's submission, revert any submissions after that with huge improvement to their previous score." }, { "id": 326502, "postDate": "2018-05-09T20:40:13.197Z", "content": "@Yunfei, I made huge improvement after that, and I did not copy or reuse any of it. Same is probably true of many participants.
\n\nThe real issue is to let people publish new kernels till the last minute. Hope this will no longer be possible in future competitions.
", "rawMarkdown": "@Yunfei, I made huge improvement after that, and I did not copy or reuse any of it. Same is probably true of many participants. \n\nThe real issue is to let people publish new kernels till the last minute. Hope this will no longer be possible in future competitions." }, { "id": 326514, "postDate": "2018-05-09T21:06:44.437Z", "rawMarkdown": "", "isDeleted": true }, { "id": 326536, "postDate": "2018-05-09T22:14:42.713Z", "content": "We can and should look forward. But if this trend exist, it will ruin kaggle with all the efforts for everyone paid on it. \nThe worst future is that when you told somebody you have a good rank on some competitions in kaggle, they said that's because you copied. And this is what Trump's criticism to China right now.
", "rawMarkdown": "We can and should look forward. But if this trend exist, it will ruin kaggle with all the efforts for everyone paid on it. \nThe worst future is that when you told somebody you have a good rank on some competitions in kaggle, they said that's because you copied. And this is what Trump's criticism to China right now." }, { "id": 326537, "postDate": "2018-05-09T22:16:02.463Z", "content": "I don't mean to disappoint you @CPMP, but I would be assured only if someone from kaggle admin team would officially announce a change in kernel policy. Otherwise this will continue.
", "rawMarkdown": "I don't mean to disappoint you @CPMP, but I would be assured only if someone from kaggle admin team would officially announce a change in kernel policy. Otherwise this will continue." }, { "id": 326539, "postDate": "2018-05-09T22:18:53.530Z", "content": "You are right, if no action is taken, next time people who are disappointed in this competition may do the same.
", "rawMarkdown": "You are right, if no action is taken, next time people who are disappointed in this competition may do the same." }, { "id": 326540, "postDate": "2018-05-09T22:21:09.543Z", "content": "@CPMP I read your post which said the improvement did not change your rank. I also remember your post after 2018 data science bowl ended which said kaggle should not encourage cheating. So I think we just have different definition on what is cheating.
", "rawMarkdown": "@CPMP I read your post which said the improvement did not change your rank. I also remember your post after 2018 data science bowl ended which said kaggle should not encourage cheating. So I think we just have different definition on what is cheating.", "votes": 1 }, { "id": 326541, "postDate": "2018-05-09T22:26:59.947Z", "content": "@Mohd I support the suggestion to forbidden to share the submission files, but only code.
", "rawMarkdown": "@Mohd I support the suggestion to forbidden to share the submission files, but only code.", "votes": 1 }, { "id": 326542, "postDate": "2018-05-09T22:27:48.840Z", "rawMarkdown": "", "isDeleted": true } ] }, { "id": 324816, "postDate": "2018-05-08T00:19:08.760Z", "content": "Well, have a look at this
", "rawMarkdown": "Well, have a look at this", "votes": 1, "replies": [ { "id": 324819, "postDate": "2018-05-08T00:21:33.383Z", "content": "There are also more than 100 people in bronze range with identical scores :/
", "rawMarkdown": "There are also more than 100 people in bronze range with identical scores :/", "votes": 3 }, { "id": 324821, "postDate": "2018-05-08T00:24:14.473Z", "content": "This one is due to the kernel, which is legit.
", "rawMarkdown": "This one is due to the kernel, which is legit.", "votes": -2 }, { "id": 324823, "postDate": "2018-05-08T00:25:49.567Z", "content": "I think the other one also counts as a kernel submission? Since the file was shared via kernel
", "rawMarkdown": "I think the other one also counts as a kernel submission? Since the file was shared via kernel" }, { "id": 324825, "postDate": "2018-05-08T00:26:42.583Z", "content": "This remember me Statoil competition...
\n\nMany people submitted a blend of blends of blends kernels and finished with silver medal above me
", "rawMarkdown": "This remember me Statoil competition...\n\nMany people submitted a blend of blends of blends kernels and finished with silver medal above me ", "votes": 2 }, { "id": 324827, "postDate": "2018-05-08T00:28:49.313Z", "content": "oh it makes use of\"df2 = pd.read_csv('../input/entire-dataset-lb-0-9811-lightgbm/sub-it200102csv')\"
", "rawMarkdown": "oh it makes use of\"df2 = pd.read_csv('../input/entire-dataset-lb-0-9811-lightgbm/sub-it200102csv')\"\n", "votes": 1 }, { "id": 324828, "postDate": "2018-05-08T00:29:01.460Z", "content": "They're not supposed to be sharing in the last week, much less the last day.
", "rawMarkdown": "They're not supposed to be sharing in the last week, much less the last day.", "votes": 1 }, { "id": 325067, "postDate": "2018-05-08T04:49:24.540Z", "rawMarkdown": "", "isDeleted": true } ] }, { "id": 325045, "postDate": "2018-05-08T03:56:53.187Z", "content": "Why don't Kaggle forbid publishing competition kernel 1 week before submission? Any objection with that?
", "rawMarkdown": "Why don't Kaggle forbid publishing competition kernel 1 week before submission? Any objection with that?", "votes": 2, "replies": [ { "id": 325085, "postDate": "2018-05-08T05:25:18.367Z", "content": "People would just share off-platform, private groups etc.. Might as well keep it out in the open where it can be monitored. (EDIT: I don't like it any more than the rest of you, and it degrades the competition experience and learning dividend)
", "rawMarkdown": "People would just share off-platform, private groups etc.. Might as well keep it out in the open where it can be monitored. (EDIT: I don't like it any more than the rest of you, and it degrades the competition experience and learning dividend)", "votes": -3 }, { "id": 325196, "postDate": "2018-05-08T08:03:15.133Z", "content": "Or they won't because it is against the rules. At least it won't be 100 people in the last day.
\n\nThat's such a bad excuse there are 10s of analogues why it isn't a bad idea.
\n\nWhy forbid stealing if some would still steal even though it is illigal?\nWhy students aren't allowed to copy papers even though some are still doing it privately?\nI can go on with more populist examples but the idea is the same.
\n\nThis time I didn't got hurt by it, but I could be hurt next time. I think this habit brings toxicity to our community.
", "rawMarkdown": "Or they won't because it is against the rules. At least it won't be 100 people in the last day.\n\nThat's such a bad excuse there are 10s of analogues why it isn't a bad idea.\n\nWhy forbid stealing if some would still steal even though it is illigal?\nWhy students aren't allowed to copy papers even though some are still doing it privately?\nI can go on with more populist examples but the idea is the same.\n\nThis time I didn't got hurt by it, but I could be hurt next time. I think this habit brings toxicity to our community.", "votes": 5 } ] }, { "id": 325441, "postDate": "2018-05-08T12:21:08.983Z", "content": "Why would it? Which rule was broken?
", "rawMarkdown": "Why would it? Which rule was broken?", "replies": [ { "id": 325475, "postDate": "2018-05-08T13:00:31.277Z", "content": "Admin requested him to keep the kernel till competition ends, but it was deleted before end time.
", "rawMarkdown": "Admin requested him to keep the kernel till competition ends, but it was deleted before end time.", "votes": 4 } ] } ], "comments": [ { "id": 325277, "author_name": "Antonis Maronikolakis", "author_url": "", "post_date": "2018-05-08T09:17:08.747000", "content": "I personally would be disappointed if people were disqualified for this. What they did may be unethical, but it is not against the rules. Retroactively changing the rules is not cool in my book.
\n\nChanging the rules after the fact happened in Data Science Bowl 2018 and people were upset (rightfully so). Changing the rules shouldn't be allowed here either. The best course of action is to change things for the better moving forward.
", "votes": 3, "replies": [] }, { "id": 325672, "author_name": "inversion", "author_url": "", "post_date": "2018-05-08T17:40:54.507000", "content": "No one will be removed for using Dirk's submission.
", "votes": 2, "replies": [ { "id": 325689, "author_name": "Snorlax", "author_url": "", "post_date": "2018-05-08T18:05:36.800000", "content": "gg
", "votes": 1, "replies": [] }, { "id": 326486, "author_name": "YUNFEI DUAN", "author_url": "", "post_date": "2018-05-09T20:04:24.637000", "content": "@inversion\nBecause of the huge data and long time for training, it is just almost no chance to get big improvement in 5 hours.\nSo I suggest that find the timestamp of Dirk's submission, revert any submissions after that with huge improvement to their previous score.
", "votes": 0, "replies": [] }, { "id": 326502, "author_name": "CPMP", "author_url": "", "post_date": "2018-05-09T20:40:13.197000", "content": "@Yunfei, I made huge improvement after that, and I did not copy or reuse any of it. Same is probably true of many participants.
\n\nThe real issue is to let people publish new kernels till the last minute. Hope this will no longer be possible in future competitions.
", "votes": 0, "replies": [] }, { "id": 326514, "author_name": "", "author_url": "", "post_date": "2018-05-09T21:06:44.437000", "content": "", "votes": 0, "replies": [] }, { "id": 326536, "author_name": "YUNFEI DUAN", "author_url": "", "post_date": "2018-05-09T22:14:42.713000", "content": "We can and should look forward. But if this trend exist, it will ruin kaggle with all the efforts for everyone paid on it. \nThe worst future is that when you told somebody you have a good rank on some competitions in kaggle, they said that's because you copied. And this is what Trump's criticism to China right now.
", "votes": 0, "replies": [] }, { "id": 326537, "author_name": "Mohammad Shadab Alam", "author_url": "", "post_date": "2018-05-09T22:16:02.463000", "content": "I don't mean to disappoint you @CPMP, but I would be assured only if someone from kaggle admin team would officially announce a change in kernel policy. Otherwise this will continue.
", "votes": 0, "replies": [] }, { "id": 326539, "author_name": "Mohammad Shadab Alam", "author_url": "", "post_date": "2018-05-09T22:18:53.530000", "content": "You are right, if no action is taken, next time people who are disappointed in this competition may do the same.
", "votes": 0, "replies": [] }, { "id": 326540, "author_name": "YUNFEI DUAN", "author_url": "", "post_date": "2018-05-09T22:21:09.543000", "content": "@CPMP I read your post which said the improvement did not change your rank. I also remember your post after 2018 data science bowl ended which said kaggle should not encourage cheating. So I think we just have different definition on what is cheating.
", "votes": 1, "replies": [] }, { "id": 326541, "author_name": "YUNFEI DUAN", "author_url": "", "post_date": "2018-05-09T22:26:59.947000", "content": "@Mohd I support the suggestion to forbidden to share the submission files, but only code.
", "votes": 1, "replies": [] }, { "id": 326542, "author_name": "", "author_url": "", "post_date": "2018-05-09T22:27:48.840000", "content": "", "votes": 0, "replies": [] } ] }, { "id": 324816, "author_name": "Snorlax", "author_url": "", "post_date": "2018-05-08T00:19:08.760000", "content": "Well, have a look at this
", "votes": 1, "replies": [ { "id": 324819, "author_name": "Joe Eddy", "author_url": "", "post_date": "2018-05-08T00:21:33.383000", "content": "There are also more than 100 people in bronze range with identical scores :/
", "votes": 3, "replies": [] }, { "id": 324821, "author_name": "Mohammad Shadab Alam", "author_url": "", "post_date": "2018-05-08T00:24:14.473000", "content": "This one is due to the kernel, which is legit.
", "votes": -2, "replies": [] }, { "id": 324823, "author_name": "Joe Eddy", "author_url": "", "post_date": "2018-05-08T00:25:49.567000", "content": "I think the other one also counts as a kernel submission? Since the file was shared via kernel
", "votes": 0, "replies": [] }, { "id": 324825, "author_name": "Serigne ", "author_url": "", "post_date": "2018-05-08T00:26:42.583000", "content": "This remember me Statoil competition...
\n\nMany people submitted a blend of blends of blends kernels and finished with silver medal above me
", "votes": 2, "replies": [] }, { "id": 324827, "author_name": "Mohammad Shadab Alam", "author_url": "", "post_date": "2018-05-08T00:28:49.313000", "content": "oh it makes use of\"df2 = pd.read_csv('../input/entire-dataset-lb-0-9811-lightgbm/sub-it200102csv')\"
", "votes": 1, "replies": [] }, { "id": 324828, "author_name": "Alan Khoa Nguyen", "author_url": "", "post_date": "2018-05-08T00:29:01.460000", "content": "They're not supposed to be sharing in the last week, much less the last day.
", "votes": 1, "replies": [] }, { "id": 325067, "author_name": "", "author_url": "", "post_date": "2018-05-08T04:49:24.540000", "content": "", "votes": 0, "replies": [] } ] }, { "id": 325045, "author_name": "Budi Ryan", "author_url": "", "post_date": "2018-05-08T03:56:53.187000", "content": "Why don't Kaggle forbid publishing competition kernel 1 week before submission? Any objection with that?
", "votes": 2, "replies": [ { "id": 325085, "author_name": "Stephen McInerney", "author_url": "", "post_date": "2018-05-08T05:25:18.367000", "content": "People would just share off-platform, private groups etc.. Might as well keep it out in the open where it can be monitored. (EDIT: I don't like it any more than the rest of you, and it degrades the competition experience and learning dividend)
", "votes": -3, "replies": [] }, { "id": 325196, "author_name": "Yair Beer", "author_url": "", "post_date": "2018-05-08T08:03:15.133000", "content": "Or they won't because it is against the rules. At least it won't be 100 people in the last day.
\n\nThat's such a bad excuse there are 10s of analogues why it isn't a bad idea.
\n\nWhy forbid stealing if some would still steal even though it is illigal?\nWhy students aren't allowed to copy papers even though some are still doing it privately?\nI can go on with more populist examples but the idea is the same.
\n\nThis time I didn't got hurt by it, but I could be hurt next time. I think this habit brings toxicity to our community.
", "votes": 5, "replies": [] } ] }, { "id": 325441, "author_name": "CPMP", "author_url": "", "post_date": "2018-05-08T12:21:08.983000", "content": "Why would it? Which rule was broken?
", "votes": 0, "replies": [ { "id": 325475, "author_name": "Mohammad Shadab Alam", "author_url": "", "post_date": "2018-05-08T13:00:31.277000", "content": "Admin requested him to keep the kernel till competition ends, but it was deleted before end time.
", "votes": 4, "replies": [] } ] } ], "raw_markdown_by_id": { "324776": "If someone selects that sub, will it count as cheating?", "325277": "I personally would be disappointed if people were disqualified for this. What they did may be unethical, but it is not against the rules. Retroactively changing the rules is not cool in my book.\n\nChanging the rules after the fact happened in [Data Science Bowl 2018](https://www.kaggle.com/c/data-science-bowl-2018/discussion/55228) and people were upset (rightfully so). Changing the rules shouldn't be allowed here either. The best course of action is to change things for the better moving forward.", "325672": "No one will be removed for using Dirk's submission. ", "324816": "Well, have a look at this", "325045": "Why don't Kaggle forbid publishing competition kernel 1 week before submission? Any objection with that?", "325441": "Why would it? Which rule was broken?" } }