Hi everyone! I would like to share my opinion about the issue we are experiencing with this competition, specifically regarding information leakage in the test data.
\nTo my knowledge, organisers made it clear from day 1 that using any type of external information other than the data provided is not allowed. In this way, I feel like we are responsible for proving that our submission is fair. In this way, all the pipeline, from data preprocessing to feature selection and model cross-validation, must be encapsulated in kaggle notebooks that can be checked by the organisers.
\nI honestly cannot share some of the opinions I have read that argue that \"they cannot just dismiss the work done participants\", I believe that if this work has not been done according to the rules of the competition, it is the responsibility of the participant and not of the organisers.
\nMoreover, I think that the organisers have a lot of work ahead of them, as their efforts should be aimed at identifying which deliveries are clean and which are not, thus truly rewarding those who have taken the competition seriously.
Hi everyone! I would like to share my opinion about the issue we are experiencing with this competition, specifically regarding information leakage in the test data.
\nTo my knowledge, organisers made it clear from day 1 that using any type of external information other than the data provided is not allowed. In this way, I feel like we are responsible for proving that our submission is fair. In this way, all the pipeline, from data preprocessing to feature selection and model cross-validation, must be encapsulated in kaggle notebooks that can be checked by the organisers.
\nI honestly cannot share some of the opinions I have read that argue that \"they cannot just dismiss the work done participants\", I believe that if this work has not been done according to the rules of the competition, it is the responsibility of the participant and not of the organisers.
\nMoreover, I think that the organisers have a lot of work ahead of them, as their efforts should be aimed at identifying which deliveries are clean and which are not, thus truly rewarding those who have taken the competition seriously.
But they said this: \"You’re welcome to use all the data we provide, as well as any freely and publicly available external datasets—including pre-trained models—to support your analysis and solution development.\" Isnt future peeking the main issue?
", "rawMarkdown": "But they said this: \"You’re welcome to use all the data we provide, as well as any freely and publicly available external datasets—including pre-trained models—to support your analysis and solution development.\" Isnt future peeking the main issue?", "votes": null }, { "id": "3249797", "postDate": "07/17/2025 05:50:20", "content": "But still, they mentioned future peeking is not allowed, as we can see the external set used by others, theoretically considered future peeking, but in general, it's just another dataset only
", "rawMarkdown": "But still, they mentioned future peeking is not allowed, as we can see the external set used by others, theoretically considered future peeking, but in general, it's just another dataset only", "votes": null }, { "id": "3249801", "postDate": "07/17/2025 06:04:41", "content": "Sorry, with test data leakage I meant future peeking. I stick to the organisers statement in the overview panel: \"We will conduct a code review to verify the reproducibility of results and ensure that no future peeking occurred during predictions. Failure to provide a compliant notebook may result in disqualification from prize eligibility.\"
", "rawMarkdown": "Sorry, with test data leakage I meant future peeking. I stick to the organisers statement in the overview panel: \"We will conduct a code review to verify the reproducibility of results and ensure that no future peeking occurred during predictions. Failure to provide a compliant notebook may result in disqualification from prize eligibility.\"", "votes": null } ], "comments": [ { "id": 3249485, "author_name": "voooolt", "author_url": "", "post_date": "07/16/2025 15:24:36", "content": "But they said this: \"You’re welcome to use all the data we provide, as well as any freely and publicly available external datasets—including pre-trained models—to support your analysis and solution development.\" Isnt future peeking the main issue?
", "votes": null, "replies": [ { "id": 3249797, "author_name": "bejjalamanisaiteja", "author_url": "", "post_date": "07/17/2025 05:50:20", "content": "But still, they mentioned future peeking is not allowed, as we can see the external set used by others, theoretically considered future peeking, but in general, it's just another dataset only
", "votes": null, "replies": [] }, { "id": 3249801, "author_name": "rafaelayllngaviln", "author_url": "", "post_date": "07/17/2025 06:04:41", "content": "Sorry, with test data leakage I meant future peeking. I stick to the organisers statement in the overview panel: \"We will conduct a code review to verify the reproducibility of results and ensure that no future peeking occurred during predictions. Failure to provide a compliant notebook may result in disqualification from prize eligibility.\"
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "3249470": "Hi everyone! I would like to share my opinion about the issue we are experiencing with this competition, specifically regarding information leakage in the test data.\nTo my knowledge, organisers made it clear from day 1 that using any type of external information other than the data provided is not allowed. In this way, I feel like **we are responsible for proving that our submission is fair. In this way, all the pipeline, from data preprocessing to feature selection and model cross-validation, must be encapsulated in kaggle notebooks that can be checked by the organisers**.\nI honestly cannot share some of the opinions I have read that argue that \"they cannot just dismiss the work done participants\", I believe that if this work has not been done according to the rules of the competition, it is the responsibility of the participant and not of the organisers.\nMoreover, I think that the organisers have a lot of work ahead of them, as their efforts should be aimed at identifying which deliveries are clean and which are not, thus truly rewarding those who have taken the competition seriously.", "3249485": "But they said this: \"You’re welcome to use all the data we provide, as well as any freely and publicly available external datasets—including pre-trained models—to support your analysis and solution development.\" Isnt future peeking the main issue?", "3249797": "But still, they mentioned future peeking is not allowed, as we can see the external set used by others, theoretically considered future peeking, but in general, it's just another dataset only", "3249801": "Sorry, with test data leakage I meant future peeking. I stick to the organisers statement in the overview panel: \"We will conduct a code review to verify the reproducibility of results and ensure that no future peeking occurred during predictions. Failure to provide a compliant notebook may result in disqualification from prize eligibility.\"" }, "source": "meta" }