Congratulations to everyone! I think we have learnt a lot from this fun competition. Thank you Amex and Kaggle hosting such unforgettable game. And Thank you Kagglers for sharing many helpful discussions and notebooks.
\nReally thanks Ivan Kireev @ivkireev who provides me a self-supervised library to FE and Med Ali Bouchhioua @medali1992 who shares the basic tabnet model. Both of them were key to get medel.
\nWhen I paid attention to feature engineering late in the competition, I was faced with the problems of no boosting and hardware. So I started to experiment with pretrain features of tabnet and the ptls library which provides self-supervision. The result is that boosting in CV and public LB.
\nIt's a little pity that in the end I did not get a feature that satisfied me, but I'will keep exploring these tow things in tabular world.
", "messages": [ { "id": 1913183, "postDate": "2022-08-25T07:52:16.340Z", "content": "Congratulations to everyone! I think we have learnt a lot from this fun competition. Thank you Amex and Kaggle hosting such unforgettable game. And Thank you Kagglers for sharing many helpful discussions and notebooks.
\nReally thanks Ivan Kireev @ivkireev who provides me a self-supervised library to FE and Med Ali Bouchhioua @medali1992 who shares the basic tabnet model. Both of them were key to get medel.
\nWhen I paid attention to feature engineering late in the competition, I was faced with the problems of no boosting and hardware. So I started to experiment with pretrain features of tabnet and the ptls library which provides self-supervision. The result is that boosting in CV and public LB.
\nIt's a little pity that in the end I did not get a feature that satisfied me, but I'will keep exploring these tow things in tabular world.
", "rawMarkdown": "Congratulations to everyone! I think we have learnt a lot from this fun competition. Thank you Amex and Kaggle hosting such unforgettable game. And Thank you Kagglers for sharing many helpful discussions and notebooks. \n\n## Some reflections on the NN-FE\nReally thanks Ivan Kireev @ivkireev who provides me a self-supervised library to FE and Med Ali Bouchhioua @medali1992 who shares the basic tabnet model. Both of them were key to get medel.\n\nWhen I paid attention to feature engineering late in the competition, I was faced with the problems of no boosting and hardware. So I started to experiment with pretrain features of tabnet and the ptls library which provides self-supervision. The result is that boosting in CV and public LB.\n\nIt's a little pity that in the end I did not get a feature that satisfied me, but I'will keep exploring these tow things in tabular world.", "votes": 12 }, { "id": 1913823, "postDate": "2022-08-25T14:25:19.317Z", "content": "Wishing you hearty congratulations for the success! Great result!!
", "rawMarkdown": "Wishing you hearty congratulations for the success! Great result!! ", "votes": 1 }, { "id": 1922191, "postDate": "2022-09-01T10:31:52.793Z", "content": "Congratulations @jacksonyou. May I invite you to participate in this survey regarding your experience on Kaggle (10 min)? This is not a scam. We are a group of researchers at the City University of Hong Kong. The survey link is: https://cityuhk.questionpro.com/survey-of-kaggle-contestants
", "rawMarkdown": "Congratulations @jacksonyou. May I invite you to participate in this survey regarding your experience on Kaggle (10 min)? This is not a scam. We are a group of researchers at the City University of Hong Kong. The survey link is: https://cityuhk.questionpro.com/survey-of-kaggle-contestants", "replies": [ { "id": 1922223, "postDate": "2022-09-01T11:02:01.050Z", "rawMarkdown": "", "isDeleted": true } ] } ], "comments": [ { "id": 1913823, "author_name": "Ravi Ramakrishnan", "author_url": "", "post_date": "2022-08-25T14:25:19.317000", "content": "Wishing you hearty congratulations for the success! Great result!!
", "votes": 1, "replies": [] }, { "id": 1922191, "author_name": "Yang Liu", "author_url": "", "post_date": "2022-09-01T10:31:52.793000", "content": "Congratulations @jacksonyou. May I invite you to participate in this survey regarding your experience on Kaggle (10 min)? This is not a scam. We are a group of researchers at the City University of Hong Kong. The survey link is: https://cityuhk.questionpro.com/survey-of-kaggle-contestants
", "votes": 0, "replies": [ { "id": 1922223, "author_name": "", "author_url": "", "post_date": "2022-09-01T11:02:01.050000", "content": "", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "1913183": "Congratulations to everyone! I think we have learnt a lot from this fun competition. Thank you Amex and Kaggle hosting such unforgettable game. And Thank you Kagglers for sharing many helpful discussions and notebooks. \n\n## Some reflections on the NN-FE\nReally thanks Ivan Kireev @ivkireev who provides me a self-supervised library to FE and Med Ali Bouchhioua @medali1992 who shares the basic tabnet model. Both of them were key to get medel.\n\nWhen I paid attention to feature engineering late in the competition, I was faced with the problems of no boosting and hardware. So I started to experiment with pretrain features of tabnet and the ptls library which provides self-supervision. The result is that boosting in CV and public LB.\n\nIt's a little pity that in the end I did not get a feature that satisfied me, but I'will keep exploring these tow things in tabular world.", "1913823": "Wishing you hearty congratulations for the success! Great result!! ", "1922191": "Congratulations @jacksonyou. May I invite you to participate in this survey regarding your experience on Kaggle (10 min)? This is not a scam. We are a group of researchers at the City University of Hong Kong. The survey link is: https://cityuhk.questionpro.com/survey-of-kaggle-contestants" } }