{ "id": 507952, "title": "Did this happen to you also?", "url": "/competitions/home-credit-credit-risk-model-stability/discussion/507952", "author_name": "Ravi Ramakrishnan", "post_date": "2024-05-28T01:06:58.336000", "votes": 21, "comment_count": 16, "views": 0, "content": "

Hello everyone,

\n

What was your best score from a submission you did not select? We could have landed in the top 60 ranks with a silver medal has we selected better. We could have got a bronze medal here had we resorted to a non-hacked solution with the same base models as the ones selected.

\n

What about you??

\n

\n

Best regards!

", "messages": [ { "id": 2840085, "postDate": "2024-05-28T01:06:58.337Z", "content": "

Hello everyone,

\n

What was your best score from a submission you did not select? We could have landed in the top 60 ranks with a silver medal has we selected better. We could have got a bronze medal here had we resorted to a non-hacked solution with the same base models as the ones selected.

\n

What about you??

\n

\n

Best regards!

", "rawMarkdown": "Hello everyone, \n\nWhat was your best score from a submission you did not select? We could have landed in the top 60 ranks with a silver medal has we selected better. We could have got a bronze medal here had we resorted to a non-hacked solution with the same base models as the ones selected. \n\nWhat about you??\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F8273630%2F5a3719b9e90ce6e707918c3decd505fc%2FOpportunityCost.png?generation=1716858321969071&alt=media)\n\nBest regards!", "votes": 19 }, { "id": 2840166, "postDate": "2024-05-28T02:04:52.820Z", "content": "

We have a 0.632 private submission. The key is that you have to hack further weeks

", "rawMarkdown": "We have a 0.632 private submission. The key is that you have to hack further weeks", "votes": 4, "replies": [ { "id": 2840483, "postDate": "2024-05-28T05:45:26.263Z", "content": "

Screenshot please

", "rawMarkdown": "Screenshot please ", "votes": -2 } ] }, { "id": 2841017, "postDate": "2024-05-28T11:05:05.660Z", "content": "

Congratulations on your gold medal!

", "rawMarkdown": "Congratulations on your gold medal! ", "votes": 1, "replies": [ { "id": 2841056, "postDate": "2024-05-28T11:31:30.637Z", "content": "

I wish I got one here @yshiraityo

", "rawMarkdown": "I wish I got one here @yshiraityo " } ] }, { "id": 2840251, "postDate": "2024-05-28T03:44:35.553Z", "content": "

Me too!!🙃
\n

", "rawMarkdown": "Me too!!🙃\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1807367%2F48a36bceb7504e6e70f8c161f277c19d%2Fimage.png?generation=1716867805776532&alt=media)", "votes": 1 }, { "id": 2848852, "postDate": "2024-06-01T08:51:08.937Z", "content": "

same here :(,
\nsilver gone, the best can land me on 150 rank with score 0.518x

", "rawMarkdown": "same here :(, \nsilver gone, the best can land me on 150 rank with score 0.518x", "votes": 2, "replies": [ { "id": 2848924, "postDate": "2024-06-01T09:24:35.907Z", "content": "

We too fell off from silver position to 1500+ rank @letemoin

", "rawMarkdown": "We too fell off from silver position to 1500+ rank @letemoin ", "votes": 1 } ] }, { "id": 2840381, "postDate": "2024-05-28T04:52:09.213Z", "content": "

It seems to have happened to everyone. It is a difficult competition to predict the best submission.

", "rawMarkdown": "It seems to have happened to everyone. It is a difficult competition to predict the best submission.", "votes": 2, "replies": [ { "id": 2841057, "postDate": "2024-05-28T11:32:07.547Z", "content": "

@yuuniekiri hearty congratulations for the win!

", "rawMarkdown": "@yuuniekiri hearty congratulations for the win!", "replies": [ { "id": 2841103, "postDate": "2024-05-28T12:05:52.390Z", "content": "

Thanks a lot. :)

", "rawMarkdown": "Thanks a lot. :)", "votes": 1 } ] } ] }, { "id": 2840263, "postDate": "2024-05-28T04:03:33.503Z", "content": "

I think one should approach from the \"Fault Tolerant Computing\" approaches of n-version programming while selecting the submissions. In this regard, I have shared my note book :
\nhttps://www.kaggle.com/code/crsuthikshnkumar/gametheory-appln-ml-competitions-crsk

", "rawMarkdown": "I think one should approach from the \"Fault Tolerant Computing\" approaches of n-version programming while selecting the submissions. In this regard, I have shared my note book :\nhttps://www.kaggle.com/code/crsuthikshnkumar/gametheory-appln-ml-competitions-crsk\n", "votes": 2 }, { "id": 2840189, "postDate": "2024-05-28T02:34:46.320Z", "content": "

Here's my case:

\n

My best solution was my public notebook with the metric hack implementation - which I did not select. Instead, I selected one that scored 0.584/0.507 to the public/private LB respectively.

", "rawMarkdown": "Here's my case:\n\nMy best solution was my public notebook with the [metric hack implementation](https://www.kaggle.com/code/andreasbis/metric-hack-implementation) - which I did not select. Instead, I selected one that scored 0.584/0.507 to the public/private LB respectively.", "votes": 2 }, { "id": 2840099, "postDate": "2024-05-28T01:19:02.567Z", "content": "

previous public score 0.608 with private 0.607, which showing running error !.

", "rawMarkdown": "previous public score 0.608 with private 0.607, which showing running error !.", "votes": 2 }, { "id": 2841086, "postDate": "2024-05-28T11:52:26.053Z", "rawMarkdown": "", "isDeleted": true }, { "id": 2840200, "postDate": "2024-05-28T02:50:02.787Z", "rawMarkdown": "", "votes": 3, "isDeleted": true }, { "id": 2840111, "postDate": "2024-05-28T01:30:49.607Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 2840166, "author_name": "minhtu.mt.mt", "author_url": "", "post_date": "2024-05-28T02:04:52.820000", "content": "

We have a 0.632 private submission. The key is that you have to hack further weeks

", "votes": 4, "replies": [ { "id": 2840483, "author_name": "JK-Piece", "author_url": "", "post_date": "2024-05-28T05:45:26.263000", "content": "

Screenshot please

", "votes": -2, "replies": [] } ] }, { "id": 2841017, "author_name": "Y. Shirai (TYO)", "author_url": "", "post_date": "2024-05-28T11:05:05.660000", "content": "

Congratulations on your gold medal!

", "votes": 1, "replies": [ { "id": 2841056, "author_name": "Ravi Ramakrishnan", "author_url": "", "post_date": "2024-05-28T11:31:30.637000", "content": "

I wish I got one here @yshiraityo

", "votes": 0, "replies": [] } ] }, { "id": 2840251, "author_name": "Isao", "author_url": "", "post_date": "2024-05-28T03:44:35.553000", "content": "

Me too!!🙃
\n

", "votes": 1, "replies": [] }, { "id": 2848852, "author_name": "Mahmoud Elshahed", "author_url": "", "post_date": "2024-06-01T08:51:08.937000", "content": "

same here :(,
\nsilver gone, the best can land me on 150 rank with score 0.518x

", "votes": 2, "replies": [ { "id": 2848924, "author_name": "Ravi Ramakrishnan", "author_url": "", "post_date": "2024-06-01T09:24:35.907000", "content": "

We too fell off from silver position to 1500+ rank @letemoin

", "votes": 1, "replies": [] } ] }, { "id": 2840381, "author_name": "yuuniee", "author_url": "", "post_date": "2024-05-28T04:52:09.213000", "content": "

It seems to have happened to everyone. It is a difficult competition to predict the best submission.

", "votes": 2, "replies": [ { "id": 2841057, "author_name": "Ravi Ramakrishnan", "author_url": "", "post_date": "2024-05-28T11:32:07.547000", "content": "

@yuuniekiri hearty congratulations for the win!

", "votes": 0, "replies": [ { "id": 2841103, "author_name": "yuuniee", "author_url": "", "post_date": "2024-05-28T12:05:52.390000", "content": "

Thanks a lot. :)

", "votes": 1, "replies": [] } ] } ] }, { "id": 2840263, "author_name": "C R Suthikshn Kumar", "author_url": "", "post_date": "2024-05-28T04:03:33.503000", "content": "

I think one should approach from the \"Fault Tolerant Computing\" approaches of n-version programming while selecting the submissions. In this regard, I have shared my note book :
\nhttps://www.kaggle.com/code/crsuthikshnkumar/gametheory-appln-ml-competitions-crsk

", "votes": 2, "replies": [] }, { "id": 2840189, "author_name": "Andreas Bisiadis", "author_url": "", "post_date": "2024-05-28T02:34:46.320000", "content": "

Here's my case:

\n

My best solution was my public notebook with the metric hack implementation - which I did not select. Instead, I selected one that scored 0.584/0.507 to the public/private LB respectively.

", "votes": 2, "replies": [] }, { "id": 2840099, "author_name": "dragon zhang", "author_url": "", "post_date": "2024-05-28T01:19:02.567000", "content": "

previous public score 0.608 with private 0.607, which showing running error !.

", "votes": 2, "replies": [] }, { "id": 2841086, "author_name": "", "author_url": "", "post_date": "2024-05-28T11:52:26.053000", "content": "", "votes": 0, "replies": [] }, { "id": 2840200, "author_name": "", "author_url": "", "post_date": "2024-05-28T02:50:02.787000", "content": "", "votes": 3, "replies": [] }, { "id": 2840111, "author_name": "", "author_url": "", "post_date": "2024-05-28T01:30:49.607000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "2840085": "Hello everyone, \n\nWhat was your best score from a submission you did not select? We could have landed in the top 60 ranks with a silver medal has we selected better. We could have got a bronze medal here had we resorted to a non-hacked solution with the same base models as the ones selected. \n\nWhat about you??\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F8273630%2F5a3719b9e90ce6e707918c3decd505fc%2FOpportunityCost.png?generation=1716858321969071&alt=media)\n\nBest regards!", "2840166": "We have a 0.632 private submission. The key is that you have to hack further weeks", "2841017": "Congratulations on your gold medal! ", "2840251": "Me too!!🙃\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1807367%2F48a36bceb7504e6e70f8c161f277c19d%2Fimage.png?generation=1716867805776532&alt=media)", "2848852": "same here :(, \nsilver gone, the best can land me on 150 rank with score 0.518x", "2840381": "It seems to have happened to everyone. It is a difficult competition to predict the best submission.", "2840263": "I think one should approach from the \"Fault Tolerant Computing\" approaches of n-version programming while selecting the submissions. In this regard, I have shared my note book :\nhttps://www.kaggle.com/code/crsuthikshnkumar/gametheory-appln-ml-competitions-crsk\n", "2840189": "Here's my case:\n\nMy best solution was my public notebook with the [metric hack implementation](https://www.kaggle.com/code/andreasbis/metric-hack-implementation) - which I did not select. Instead, I selected one that scored 0.584/0.507 to the public/private LB respectively.", "2840099": "previous public score 0.608 with private 0.607, which showing running error !.", "2841086": "", "2840200": "", "2840111": "" } }