{ "id": 306175, "title": "Clarification of Rule: Is Accessing submission.csv is Officially Prohibited?", "url": "/competitions/tensorflow-great-barrier-reef/discussion/306175", "author_name": "", "post_date": "2022-02-08T12:24:58.728327300Z", "votes": 4, "comment_count": 6, "views": 0, "content": "

@addisonhoward
\n@sohier

\n

Dear Kaggle staff,

\n

I would like to clarify the competition rules.
\nIs it officially forbidden to access submission.csv?
\nIn other words, if someone reads from submission.csv or writes to submission.csv, does that disqualify them from the competition?

\n

As I wrote in the following discussion [1], people who know the distribution of the background frames have an advantage over others.

\n

On the other hand, by directly editing submission.csv, we can possibly know about the distribution of the background frames of hidden test set.

\n

So, I would like to clarify whether editing or viewing submission.csv is forbidden as a competition rule or not.

\n", "messages": [ { "id": "1681341", "postDate": "02/08/2022 12:24:58", "content": "

@addisonhoward
\n@sohier

\n

Dear Kaggle staff,

\n

I would like to clarify the competition rules.
\nIs it officially forbidden to access submission.csv?
\nIn other words, if someone reads from submission.csv or writes to submission.csv, does that disqualify them from the competition?

\n

As I wrote in the following discussion [1], people who know the distribution of the background frames have an advantage over others.

\n

On the other hand, by directly editing submission.csv, we can possibly know about the distribution of the background frames of hidden test set.

\n

So, I would like to clarify whether editing or viewing submission.csv is forbidden as a competition rule or not.

\n", "rawMarkdown": "addisonhoward\n@sohier \n\nDear Kaggle staff,\n\nI would like to clarify the competition rules.\nIs it officially forbidden to access `submission.csv`?\nIn other words, if someone reads from `submission.csv` or writes to `submission.csv`, does that disqualify them from the competition?\n\nAs I wrote in the following discussion [1], people who know the distribution of the background frames have an advantage over others.\n\nOn the other hand, by directly editing `submission.csv`, we can possibly know about the distribution of the background frames of hidden test set.\n\nSo, I would like to clarify whether editing or viewing `submission.csv` is forbidden as a competition rule or not.\n\n* [1] https://www.kaggle.com/c/tensorflow-great-barrier-reef/discussion/306143", "votes": null }, { "id": "1681520", "postDate": "02/08/2022 14:23:12", "content": "

best models won't work like this way.

", "rawMarkdown": "best models won't work like this way.", "votes": null }, { "id": "1681600", "postDate": "02/08/2022 15:21:48", "content": "

The time-series API does not allow accessing/modifying the submission file directly.

", "rawMarkdown": "The time-series API does not allow accessing/modifying the submission file directly.", "votes": null }, { "id": "1681626", "postDate": "02/08/2022 15:37:28", "content": "

@wcukierski Thank you for clarification.

\n

Does that mean we don’t have read/write permission to submission.csv?
\nI wonder how this is technically achieved.

", "rawMarkdown": "wcukierski Thank you for clarification.\n\nDoes that mean we don’t have read/write permission to submission.csv?\nI wonder how this is technically achieved.", "votes": null }, { "id": "1681651", "postDate": "02/08/2022 15:58:19", "content": "

You can think of it that way. We don't share the technical details for obvious reasons :)

", "rawMarkdown": "You can think of it that way. We don't share the technical details for obvious reasons :)", "votes": null }, { "id": "1681653", "postDate": "02/08/2022 15:59:06", "content": "

I don’t say some of top ranker’s are doing this kind of LB probing.
\nI just want to clear out the possibility of unfair LB probing.

", "rawMarkdown": "I don’t say some of top ranker’s are doing this kind of LB probing.\nI just want to clear out the possibility of unfair LB probing.", "votes": null }, { "id": "1681655", "postDate": "02/08/2022 15:59:54", "content": "

I see. Thanks anyway.

", "rawMarkdown": "I see. Thanks anyway.", "votes": null } ], "comments": [ { "id": 1681520, "author_name": "dragonzhang", "author_url": "", "post_date": "02/08/2022 14:23:12", "content": "

best models won't work like this way.

", "votes": null, "replies": [ { "id": 1681653, "author_name": "tatamikenn", "author_url": "", "post_date": "02/08/2022 15:59:06", "content": "

I don’t say some of top ranker’s are doing this kind of LB probing.
\nI just want to clear out the possibility of unfair LB probing.

", "votes": null, "replies": [] } ] }, { "id": 1681600, "author_name": "wcukierski", "author_url": "", "post_date": "02/08/2022 15:21:48", "content": "

The time-series API does not allow accessing/modifying the submission file directly.

", "votes": null, "replies": [ { "id": 1681626, "author_name": "tatamikenn", "author_url": "", "post_date": "02/08/2022 15:37:28", "content": "

@wcukierski Thank you for clarification.

\n

Does that mean we don’t have read/write permission to submission.csv?
\nI wonder how this is technically achieved.

", "votes": null, "replies": [] }, { "id": 1681651, "author_name": "wcukierski", "author_url": "", "post_date": "02/08/2022 15:58:19", "content": "

You can think of it that way. We don't share the technical details for obvious reasons :)

", "votes": null, "replies": [] }, { "id": 1681655, "author_name": "tatamikenn", "author_url": "", "post_date": "02/08/2022 15:59:54", "content": "

I see. Thanks anyway.

", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1681341": "addisonhoward\n@sohier \n\nDear Kaggle staff,\n\nI would like to clarify the competition rules.\nIs it officially forbidden to access `submission.csv`?\nIn other words, if someone reads from `submission.csv` or writes to `submission.csv`, does that disqualify them from the competition?\n\nAs I wrote in the following discussion [1], people who know the distribution of the background frames have an advantage over others.\n\nOn the other hand, by directly editing `submission.csv`, we can possibly know about the distribution of the background frames of hidden test set.\n\nSo, I would like to clarify whether editing or viewing `submission.csv` is forbidden as a competition rule or not.\n\n* [1] https://www.kaggle.com/c/tensorflow-great-barrier-reef/discussion/306143", "1681520": "best models won't work like this way.", "1681600": "The time-series API does not allow accessing/modifying the submission file directly.", "1681626": "wcukierski Thank you for clarification.\n\nDoes that mean we don’t have read/write permission to submission.csv?\nI wonder how this is technically achieved.", "1681651": "You can think of it that way. We don't share the technical details for obvious reasons :)", "1681653": "I don’t say some of top ranker’s are doing this kind of LB probing.\nI just want to clear out the possibility of unfair LB probing.", "1681655": "I see. Thanks anyway." }, "source": "meta" }