Dear respected Kaggle (grand)masters.
\nSorry to bother you, but please give me a second.
May I kindly attract your attention to that particular competition, and ask you to consider the possibility of joining to it.
\nIn my mind that competition is kind of exceptional on Kaggle, because the data considered here is new, but in forthcoming years dozens similar datasets will be produced and later hundreds. So insights which you may find - can be of DIRECT use for many researchers and potentially in future may be useful for cancer immunotherapy and, so saving people's lives.
\nAlthough improvement of score on tiny epsilon is not useful for practice, but the ideas which can be found during that, might be useful.
The competition is 26 days to go - so there is some time.
\nIt is more or less classical tabular data, but multitarget. Actually two datasets are given and two models should be made.
\nThe score is like average (weighted by sample size) on the two.
Current public approaches and last year same competition shows that up to now only simple ideas work.
\n(PCA + selected features) + simple neural network (4 fully connected layers and dropout) + CatBoost.
\nI mean for (grand)masters there is no magic here and getting medal I guess is not a big problem for you.
Some other ideas are collected here:
\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/348311
\nand here:
\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/361025
If any questions - would be happy to try to answer, and/or ask my colleagues to provide more biological background.
\nI think any your contribution would be valuable and, one may consider to make a publication out it, or we can try to make it jointly with biologists.
\nSincerely Yours,
\nAlexander
@christofhenkel @bestfitting @philippsinger @wowfattie @khyeh0719 @titericz @haqishen @ilu000 @cpmpml @tascj0 @cdeotte @nvnnghia @selimsef @aerdem4 @vityzi @charmq @takuok @nischaydnk @daishu @bo6260463 @coreacasa @kirderf @taroz1461 @steamedsheep @theoviel @takoihiraokazu @tikutiku @ybabakhin @goldenlock
", "messages": [ { "id": "1997861", "postDate": "10/21/2022 07:50:34", "content": "Dear respected Kaggle (grand)masters.
\nSorry to bother you, but please give me a second.
May I kindly attract your attention to that particular competition, and ask you to consider the possibility of joining to it.
\nIn my mind that competition is kind of exceptional on Kaggle, because the data considered here is new, but in forthcoming years dozens similar datasets will be produced and later hundreds. So insights which you may find - can be of DIRECT use for many researchers and potentially in future may be useful for cancer immunotherapy and, so saving people's lives.
\nAlthough improvement of score on tiny epsilon is not useful for practice, but the ideas which can be found during that, might be useful.
The competition is 26 days to go - so there is some time.
\nIt is more or less classical tabular data, but multitarget. Actually two datasets are given and two models should be made.
\nThe score is like average (weighted by sample size) on the two.
Current public approaches and last year same competition shows that up to now only simple ideas work.
\n(PCA + selected features) + simple neural network (4 fully connected layers and dropout) + CatBoost.
\nI mean for (grand)masters there is no magic here and getting medal I guess is not a big problem for you.
Some other ideas are collected here:
\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/348311
\nand here:
\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/361025
If any questions - would be happy to try to answer, and/or ask my colleagues to provide more biological background.
\nI think any your contribution would be valuable and, one may consider to make a publication out it, or we can try to make it jointly with biologists.
\nSincerely Yours,
\nAlexander
@christofhenkel @bestfitting @philippsinger @wowfattie @khyeh0719 @titericz @haqishen @ilu000 @cpmpml @tascj0 @cdeotte @nvnnghia @selimsef @aerdem4 @vityzi @charmq @takuok @nischaydnk @daishu @bo6260463 @coreacasa @kirderf @taroz1461 @steamedsheep @theoviel @takoihiraokazu @tikutiku @ybabakhin @goldenlock
", "rawMarkdown": "Dear respected Kaggle (grand)masters.\nSorry to bother you, but please give me a second. \n\nMay I kindly attract your attention to that particular competition, and ask you to consider the possibility of joining to it.\nIn my mind that competition is kind of exceptional on Kaggle, because the data considered here is new, but in forthcoming years dozens similar datasets will be produced and later hundreds. So insights which you may find - can be of DIRECT use for many researchers and potentially in future may be useful for cancer immunotherapy and, so saving people's lives.\nAlthough improvement of score on tiny epsilon is not useful for practice, but the ideas which can be found during that, might be useful.\n\nThe competition is 26 days to go - so there is some time.\nIt is more or less classical tabular data, but multitarget. Actually two datasets are given and two models should be made.\nThe score is like average (weighted by sample size) on the two. \n\nCurrent public approaches and last year same competition shows that up to now only simple ideas work.\n(PCA + selected features) + simple neural network (4 fully connected layers and dropout) + CatBoost.\nI mean for (grand)masters there is no magic here and getting medal I guess is not a big problem for you.\n\nSome other ideas are collected here:\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/348311\nand here:\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/361025\n\nIf any questions - would be happy to try to answer, and/or ask my colleagues to provide more biological background.\n\nI think any your contribution would be valuable and, one may consider to make a publication out it, or we can try to make it jointly with biologists.\n\nSincerely Yours,\nAlexander\n\n\n@christofhenkel @bestfitting @philippsinger @wowfattie @khyeh0719 @titericz @haqishen @ilu000 @cpmpml @tascj0 @cdeotte @nvnnghia @selimsef @aerdem4 @vityzi @charmq @takuok @nischaydnk @daishu @bo6260463 @coreacasa @kirderf @taroz1461 @steamedsheep @theoviel @takoihiraokazu @tikutiku @ybabakhin @goldenlock", "votes": null }, { "id": "1997906", "postDate": "10/21/2022 08:17:36", "content": "Perhaps some will join after RSNA ;)
", "rawMarkdown": "Perhaps some will join after RSNA ;)", "votes": null }, { "id": "1998147", "postDate": "10/21/2022 12:16:27", "content": "I don't know for others, but I've been busy outside Kaggle since August. I'll look at Kaggle again in November.
", "rawMarkdown": "I don't know for others, but I've been busy outside Kaggle since August. I'll look at Kaggle again in November.", "votes": null }, { "id": "1998449", "postDate": "10/21/2022 16:04:58", "content": "Trying to handle them in some kind of order ;)
", "rawMarkdown": "Trying to handle them in some kind of order ;)", "votes": null }, { "id": "1999935", "postDate": "10/22/2022 21:25:09", "content": "Hi A. Chervov,
\nyou're an amazing professional with great background on that subject. Totally agree, this competiton is unique and deserves to get more kagglers attention.
\nSometimes we have beginners that could add high-quality to that research: like sidneytiagosilva (September winner kaggle prize). I prefer not to add the @ commenting his name. Maybe I'll leave a comment on his work.
\nBy the way, those that are really \"Saving the World\" aren't on the same page, or even on the same book that those that are longing for medals/prizes and more recognition. I don't even know if those saving the world have time to be on Kaggle.
\nKaggle need Grand Masters and so do they too. Mostly Competition GMs. That symbiosis is a two way street.
\nI understand those that are with short time. Many times we do what we can. Not what we want.
\nVery likely I will get many disapproval. As you got too. Who cares?
\nThe whole world is full of angry birds. Some are extremely talented, others just angry.
\nBest regards,
\nMarília.
", "rawMarkdown": "Hi A. Chervov,\n\nyou're an amazing professional with great background on that subject. Totally agree, this competiton is unique and deserves to get more kagglers attention. \n\nSometimes we have beginners that could add high-quality to that research: like sidneytiagosilva (September winner kaggle prize). I prefer not to add the @ commenting his name. Maybe I'll leave a comment on his work.\n\nBy the way, those that are really \"Saving the World\" aren't on the same page, or even on the same book that those that are longing for medals/prizes and more recognition. I don't even know if those saving the world have time to be on Kaggle. \n\nKaggle need Grand Masters and so do they too. Mostly Competition GMs. That symbiosis is a two way street.\n\nI understand those that are with short time. Many times we do what we can. Not what we want. \n\nVery likely I will get many disapproval. As you got too. Who cares?\n\nThe whole world is full of angry birds. Some are extremely talented, others just angry.\n\nBest regards,\n\nMarília.", "votes": null }, { "id": "1999975", "postDate": "10/22/2022 23:18:08", "content": "A downvote for responding to a post that tagged me.
\nI see that Kaggle community still has its haters.
", "rawMarkdown": "A downvote for responding to a post that tagged me.\n\nI see that Kaggle community still has its haters.", "votes": null }, { "id": "2000691", "postDate": "10/23/2022 13:43:40", "content": "Many thanks Marília for your kind words !
\nAnd thanks for \"Totally agree, this competiton is unique and deserves to get more kagglers attention.\"
\nHope some will hear us !
Unfortunately, Sidney isn't available. He has an end-of-term job at the university.
\nMaybe, I will write a topic about the unicity of this competition to see if it can attract more kagglers to join it.
", "rawMarkdown": "Unfortunately, Sidney isn't available. He has an end-of-term job at the university. \n\nMaybe, I will write a topic about the unicity of this competition to see if it can attract more kagglers to join it.", "votes": null }, { "id": "2001446", "postDate": "10/24/2022 03:55:52", "content": "I remember the Covid competitions with no prize money, contributors probably didnt save the world, I think they are heroes.
", "rawMarkdown": "I remember the Covid competitions with no prize money, contributors probably didnt save the world, I think they are heroes.", "votes": null }, { "id": "2002222", "postDate": "10/24/2022 16:07:55", "content": "I don't get why there are so many dislike to a post like this one… @alexandervc is really active and he is investing such a lot of time and effort in publishing high informative notebooks for this comp. Let's make kaggle a platform where we can share, grow and learn together even if sometimes this requires to poke the best guys around 🙌
", "rawMarkdown": "I don't get why there are so many dislike to a post like this one... @alexandervc is really active and he is investing such a lot of time and effort in publishing high informative notebooks for this comp. Let's make kaggle a platform where we can share, grow and learn together even if sometimes this requires to poke the best guys around 🙌", "votes": null }, { "id": "2002471", "postDate": "10/24/2022 20:39:15", "content": "Indeed, DefinitelynotB.
\nIf we don't like something we should ignore it. Dislike is a waste of time, mostly for professionals.
\nAlgorithms need movement, which means: they are attracted by many votes and many dislikes too.
What doesn't work with Algorithms: ignoring posts (zero votes) and topics with no comments. Therefore, the more comments , more views too.
\nAs you have mentioned, Chervov is very active, doing his best publishing code and topics. Which is a disrespectful attitude with an user that was just trying to call attention for the importance of this competition for research. Now and in the future.
\nBy the way, Dataclown at the beach?? Ma non c' è spiaggia in Milano : )
", "rawMarkdown": "Indeed, DefinitelynotB. \n\nIf we don't like something we should ignore it. Dislike is a waste of time, mostly for professionals.\nAlgorithms need movement, which means: they are attracted by many votes and many dislikes too. \n\nWhat doesn't work with Algorithms: ignoring posts (zero votes) and topics with no comments. Therefore, the more comments , more views too.\n\nAs you have mentioned, Chervov is very active, doing his best publishing code and topics. Which is a disrespectful attitude with an user that was just trying to call attention for the importance of this competition for research. Now and in the future.\n\nBy the way, Dataclown at the beach?? Ma non c' è spiaggia in Milano : )", "votes": null }, { "id": "2002520", "postDate": "10/24/2022 21:43:35", "content": "That's my impression as well…hope I still be able to win a medal by then 😂
", "rawMarkdown": "That's my impression as well...hope I still be able to win a medal by then 😂", "votes": null }, { "id": "2002653", "postDate": "10/25/2022 01:30:30", "content": "Since this is a competitive event, there are some innovations that we have not told you about. However, we would not be happy if you win a medal in a competition with few Grand Master participants, so please play against us.I have the utmost respect for Alexander Cherubov.
\n対戦イベントなので、お伝えしていない工夫もあります。ただ、グランドマスターの参加者が少ない大会でメダルを取っても嬉しくないので、ぜひ対戦をお願いします。私は、アレクサンダー・チェルボフを心から尊敬します。
", "rawMarkdown": "Since this is a competitive event, there are some innovations that we have not told you about. However, we would not be happy if you win a medal in a competition with few Grand Master participants, so please play against us.I have the utmost respect for Alexander Cherubov.\n\n対戦イベントなので、お伝えしていない工夫もあります。ただ、グランドマスターの参加者が少ない大会でメダルを取っても嬉しくないので、ぜひ対戦をお願いします。私は、アレクサンダー・チェルボフを心から尊敬します。", "votes": null }, { "id": "2002932", "postDate": "10/25/2022 06:56:15", "content": "That's why we like remote working ;) ☀️
", "rawMarkdown": "That's why we like remote working ;) ☀️", "votes": null }, { "id": "2028651", "postDate": "11/14/2022 06:13:59", "content": "Thank you very much, it's very useful :D
", "rawMarkdown": "Thank you very much, it's very useful :D", "votes": null }, { "id": "2028662", "postDate": "11/14/2022 06:37:18", "content": "Thank you! This is great!
", "rawMarkdown": "Thank you! This is great!", "votes": null }, { "id": "2028710", "postDate": "11/14/2022 07:32:39", "content": "Excellent!Thanks
", "rawMarkdown": "Excellent!Thanks", "votes": null } ], "comments": [ { "id": 1997906, "author_name": "theoviel", "author_url": "", "post_date": "10/21/2022 08:17:36", "content": "Perhaps some will join after RSNA ;)
", "votes": null, "replies": [ { "id": 2002520, "author_name": "leandrodestefani", "author_url": "", "post_date": "10/24/2022 21:43:35", "content": "That's my impression as well…hope I still be able to win a medal by then 😂
", "votes": null, "replies": [] } ] }, { "id": 1998147, "author_name": "cpmpml", "author_url": "", "post_date": "10/21/2022 12:16:27", "content": "I don't know for others, but I've been busy outside Kaggle since August. I'll look at Kaggle again in November.
", "votes": null, "replies": [ { "id": 1999975, "author_name": "cpmpml", "author_url": "", "post_date": "10/22/2022 23:18:08", "content": "A downvote for responding to a post that tagged me.
\nI see that Kaggle community still has its haters.
", "votes": null, "replies": [] } ] }, { "id": 1998449, "author_name": "kirderf", "author_url": "", "post_date": "10/21/2022 16:04:58", "content": "Trying to handle them in some kind of order ;)
", "votes": null, "replies": [] }, { "id": 1999935, "author_name": "mpwolke", "author_url": "", "post_date": "10/22/2022 21:25:09", "content": "Hi A. Chervov,
\nyou're an amazing professional with great background on that subject. Totally agree, this competiton is unique and deserves to get more kagglers attention.
\nSometimes we have beginners that could add high-quality to that research: like sidneytiagosilva (September winner kaggle prize). I prefer not to add the @ commenting his name. Maybe I'll leave a comment on his work.
\nBy the way, those that are really \"Saving the World\" aren't on the same page, or even on the same book that those that are longing for medals/prizes and more recognition. I don't even know if those saving the world have time to be on Kaggle.
\nKaggle need Grand Masters and so do they too. Mostly Competition GMs. That symbiosis is a two way street.
\nI understand those that are with short time. Many times we do what we can. Not what we want.
\nVery likely I will get many disapproval. As you got too. Who cares?
\nThe whole world is full of angry birds. Some are extremely talented, others just angry.
\nBest regards,
\nMarília.
", "votes": null, "replies": [ { "id": 2000691, "author_name": "alexandervc", "author_url": "", "post_date": "10/23/2022 13:43:40", "content": "Many thanks Marília for your kind words !
\nAnd thanks for \"Totally agree, this competiton is unique and deserves to get more kagglers attention.\"
\nHope some will hear us !
Unfortunately, Sidney isn't available. He has an end-of-term job at the university.
\nMaybe, I will write a topic about the unicity of this competition to see if it can attract more kagglers to join it.
", "votes": null, "replies": [] }, { "id": 2001446, "author_name": "frankw", "author_url": "", "post_date": "10/24/2022 03:55:52", "content": "I remember the Covid competitions with no prize money, contributors probably didnt save the world, I think they are heroes.
", "votes": null, "replies": [] } ] }, { "id": 2002222, "author_name": "callmeb", "author_url": "", "post_date": "10/24/2022 16:07:55", "content": "I don't get why there are so many dislike to a post like this one… @alexandervc is really active and he is investing such a lot of time and effort in publishing high informative notebooks for this comp. Let's make kaggle a platform where we can share, grow and learn together even if sometimes this requires to poke the best guys around 🙌
", "votes": null, "replies": [ { "id": 2002471, "author_name": "mpwolke", "author_url": "", "post_date": "10/24/2022 20:39:15", "content": "Indeed, DefinitelynotB.
\nIf we don't like something we should ignore it. Dislike is a waste of time, mostly for professionals.
\nAlgorithms need movement, which means: they are attracted by many votes and many dislikes too.
What doesn't work with Algorithms: ignoring posts (zero votes) and topics with no comments. Therefore, the more comments , more views too.
\nAs you have mentioned, Chervov is very active, doing his best publishing code and topics. Which is a disrespectful attitude with an user that was just trying to call attention for the importance of this competition for research. Now and in the future.
\nBy the way, Dataclown at the beach?? Ma non c' è spiaggia in Milano : )
", "votes": null, "replies": [] }, { "id": 2002932, "author_name": "callmeb", "author_url": "", "post_date": "10/25/2022 06:56:15", "content": "That's why we like remote working ;) ☀️
", "votes": null, "replies": [] } ] }, { "id": 2002653, "author_name": "miyawakiyoshifumi", "author_url": "", "post_date": "10/25/2022 01:30:30", "content": "Since this is a competitive event, there are some innovations that we have not told you about. However, we would not be happy if you win a medal in a competition with few Grand Master participants, so please play against us.I have the utmost respect for Alexander Cherubov.
\n対戦イベントなので、お伝えしていない工夫もあります。ただ、グランドマスターの参加者が少ない大会でメダルを取っても嬉しくないので、ぜひ対戦をお願いします。私は、アレクサンダー・チェルボフを心から尊敬します。
", "votes": null, "replies": [] }, { "id": 2028651, "author_name": "hermanshow", "author_url": "", "post_date": "11/14/2022 06:13:59", "content": "Thank you very much, it's very useful :D
", "votes": null, "replies": [] }, { "id": 2028662, "author_name": "", "author_url": "", "post_date": "11/14/2022 06:37:18", "content": "Thank you! This is great!
", "votes": null, "replies": [] }, { "id": 2028710, "author_name": "", "author_url": "", "post_date": "11/14/2022 07:32:39", "content": "Excellent!Thanks
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1997861": "Dear respected Kaggle (grand)masters.\nSorry to bother you, but please give me a second. \n\nMay I kindly attract your attention to that particular competition, and ask you to consider the possibility of joining to it.\nIn my mind that competition is kind of exceptional on Kaggle, because the data considered here is new, but in forthcoming years dozens similar datasets will be produced and later hundreds. So insights which you may find - can be of DIRECT use for many researchers and potentially in future may be useful for cancer immunotherapy and, so saving people's lives.\nAlthough improvement of score on tiny epsilon is not useful for practice, but the ideas which can be found during that, might be useful.\n\nThe competition is 26 days to go - so there is some time.\nIt is more or less classical tabular data, but multitarget. Actually two datasets are given and two models should be made.\nThe score is like average (weighted by sample size) on the two. \n\nCurrent public approaches and last year same competition shows that up to now only simple ideas work.\n(PCA + selected features) + simple neural network (4 fully connected layers and dropout) + CatBoost.\nI mean for (grand)masters there is no magic here and getting medal I guess is not a big problem for you.\n\nSome other ideas are collected here:\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/348311\nand here:\nhttps://www.kaggle.com/competitions/open-problems-multimodal/discussion/361025\n\nIf any questions - would be happy to try to answer, and/or ask my colleagues to provide more biological background.\n\nI think any your contribution would be valuable and, one may consider to make a publication out it, or we can try to make it jointly with biologists.\n\nSincerely Yours,\nAlexander\n\n\n@christofhenkel @bestfitting @philippsinger @wowfattie @khyeh0719 @titericz @haqishen @ilu000 @cpmpml @tascj0 @cdeotte @nvnnghia @selimsef @aerdem4 @vityzi @charmq @takuok @nischaydnk @daishu @bo6260463 @coreacasa @kirderf @taroz1461 @steamedsheep @theoviel @takoihiraokazu @tikutiku @ybabakhin @goldenlock", "1997906": "Perhaps some will join after RSNA ;)", "1998147": "I don't know for others, but I've been busy outside Kaggle since August. I'll look at Kaggle again in November.", "1998449": "Trying to handle them in some kind of order ;)", "1999935": "Hi A. Chervov,\n\nyou're an amazing professional with great background on that subject. Totally agree, this competiton is unique and deserves to get more kagglers attention. \n\nSometimes we have beginners that could add high-quality to that research: like sidneytiagosilva (September winner kaggle prize). I prefer not to add the @ commenting his name. Maybe I'll leave a comment on his work.\n\nBy the way, those that are really \"Saving the World\" aren't on the same page, or even on the same book that those that are longing for medals/prizes and more recognition. I don't even know if those saving the world have time to be on Kaggle. \n\nKaggle need Grand Masters and so do they too. Mostly Competition GMs. That symbiosis is a two way street.\n\nI understand those that are with short time. Many times we do what we can. Not what we want. \n\nVery likely I will get many disapproval. As you got too. Who cares?\n\nThe whole world is full of angry birds. Some are extremely talented, others just angry.\n\nBest regards,\n\nMarília.", "1999975": "A downvote for responding to a post that tagged me.\n\nI see that Kaggle community still has its haters.", "2000691": "Many thanks Marília for your kind words ! \nAnd thanks for \"Totally agree, this competiton is unique and deserves to get more kagglers attention.\"\nHope some will hear us !", "2001283": "Unfortunately, Sidney isn't available. He has an end-of-term job at the university. \n\nMaybe, I will write a topic about the unicity of this competition to see if it can attract more kagglers to join it.", "2001446": "I remember the Covid competitions with no prize money, contributors probably didnt save the world, I think they are heroes.", "2002222": "I don't get why there are so many dislike to a post like this one... @alexandervc is really active and he is investing such a lot of time and effort in publishing high informative notebooks for this comp. Let's make kaggle a platform where we can share, grow and learn together even if sometimes this requires to poke the best guys around 🙌", "2002471": "Indeed, DefinitelynotB. \n\nIf we don't like something we should ignore it. Dislike is a waste of time, mostly for professionals.\nAlgorithms need movement, which means: they are attracted by many votes and many dislikes too. \n\nWhat doesn't work with Algorithms: ignoring posts (zero votes) and topics with no comments. Therefore, the more comments , more views too.\n\nAs you have mentioned, Chervov is very active, doing his best publishing code and topics. Which is a disrespectful attitude with an user that was just trying to call attention for the importance of this competition for research. Now and in the future.\n\nBy the way, Dataclown at the beach?? Ma non c' è spiaggia in Milano : )", "2002520": "That's my impression as well...hope I still be able to win a medal by then 😂", "2002653": "Since this is a competitive event, there are some innovations that we have not told you about. However, we would not be happy if you win a medal in a competition with few Grand Master participants, so please play against us.I have the utmost respect for Alexander Cherubov.\n\n対戦イベントなので、お伝えしていない工夫もあります。ただ、グランドマスターの参加者が少ない大会でメダルを取っても嬉しくないので、ぜひ対戦をお願いします。私は、アレクサンダー・チェルボフを心から尊敬します。", "2002932": "That's why we like remote working ;) ☀️", "2028651": "Thank you very much, it's very useful :D", "2028662": "Thank you! This is great!", "2028710": "Excellent!Thanks" }, "source": "meta" }