There is target leakage (at least in the public leaderboard).
\nI got a perfect score without training any model.
", "messages": [ { "id": "3417346", "postDate": "03/05/2026 06:35:29", "content": "There is target leakage (at least in the public leaderboard).
\nI got a perfect score without training any model.
", "rawMarkdown": "There is target leakage (at least in the public leaderboard). \n\nI got a perfect score without training any model.", "votes": null }, { "id": "3417353", "postDate": "03/05/2026 07:14:59", "content": "@dersecure Please pay attention to this issue. There is a notebook here pointing out an ID leak.
\nAdditionally, I suspect there might be other data leaks. Even without using the ID, CatBoost easily reaches a score of 1.0 (however, a decision tree cannot. This indicates that the data leakage might not be as obvious as the ID leak).
", "rawMarkdown": "dersecure Please pay attention to this issue. There is a [notebook](https://www.kaggle.com/code/alvaroborras/f2-1-0-in-3-lines-of-code) here pointing out an ID leak. \n\nAdditionally, I suspect there might be other data leaks. Even without using the ID, CatBoost easily reaches a score of 1.0 (however, a decision tree cannot. This indicates that the data leakage might not be as obvious as the ID leak).", "votes": null }, { "id": "3417476", "postDate": "03/05/2026 13:42:46", "content": "Sorry about the dataset issues, we are working on fixing it now. The way we generate this data is very new so we expected to run into issues so we are grateful for your comments and we will keep fixing any issues that come up and sorry again for any inconvenience.
", "rawMarkdown": "Sorry about the dataset issues, we are working on fixing it now. The way we generate this data is very new so we expected to run into issues so we are grateful for your comments and we will keep fixing any issues that come up and sorry again for any inconvenience.", "votes": null }, { "id": "3417512", "postDate": "03/05/2026 14:43:46", "content": "It's understandable, and thank you very much for addressing these issues. Let us know when the new version of the dataset is updated, so that we can re-download it.
", "rawMarkdown": "It's understandable, and thank you very much for addressing these issues. Let us know when the new version of the dataset is updated, so that we can re-download it.", "votes": null }, { "id": "3417542", "postDate": "03/05/2026 16:30:43", "content": "@act18l @alpharh0 https://www.kaggle.com/competitions/cyber-physical-anomaly-detection-for-der-systems/discussion/680068
", "rawMarkdown": "act18l @alpharh0 https://www.kaggle.com/competitions/cyber-physical-anomaly-detection-for-der-systems/discussion/680068", "votes": null } ], "comments": [ { "id": 3417353, "author_name": "act18l", "author_url": "", "post_date": "03/05/2026 07:14:59", "content": "@dersecure Please pay attention to this issue. There is a notebook here pointing out an ID leak.
\nAdditionally, I suspect there might be other data leaks. Even without using the ID, CatBoost easily reaches a score of 1.0 (however, a decision tree cannot. This indicates that the data leakage might not be as obvious as the ID leak).
", "votes": null, "replies": [] }, { "id": 3417476, "author_name": "jpinedaa", "author_url": "", "post_date": "03/05/2026 13:42:46", "content": "Sorry about the dataset issues, we are working on fixing it now. The way we generate this data is very new so we expected to run into issues so we are grateful for your comments and we will keep fixing any issues that come up and sorry again for any inconvenience.
", "votes": null, "replies": [ { "id": 3417512, "author_name": "alpharh0", "author_url": "", "post_date": "03/05/2026 14:43:46", "content": "It's understandable, and thank you very much for addressing these issues. Let us know when the new version of the dataset is updated, so that we can re-download it.
", "votes": null, "replies": [] } ] }, { "id": 3417542, "author_name": "jpinedaa", "author_url": "", "post_date": "03/05/2026 16:30:43", "content": "@act18l @alpharh0 https://www.kaggle.com/competitions/cyber-physical-anomaly-detection-for-der-systems/discussion/680068
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "3417346": "There is target leakage (at least in the public leaderboard). \n\nI got a perfect score without training any model.", "3417353": "dersecure Please pay attention to this issue. There is a [notebook](https://www.kaggle.com/code/alvaroborras/f2-1-0-in-3-lines-of-code) here pointing out an ID leak. \n\nAdditionally, I suspect there might be other data leaks. Even without using the ID, CatBoost easily reaches a score of 1.0 (however, a decision tree cannot. This indicates that the data leakage might not be as obvious as the ID leak).", "3417476": "Sorry about the dataset issues, we are working on fixing it now. The way we generate this data is very new so we expected to run into issues so we are grateful for your comments and we will keep fixing any issues that come up and sorry again for any inconvenience.", "3417512": "It's understandable, and thank you very much for addressing these issues. Let us know when the new version of the dataset is updated, so that we can re-download it.", "3417542": "act18l @alpharh0 https://www.kaggle.com/competitions/cyber-physical-anomaly-detection-for-der-systems/discussion/680068" }, "source": "meta" }