I just realized my team lost positions in a week (6 positions down today). As a result of posting a high scoring kernel after the deadline, it kinda pisses me off, but i have seen this a lot more times in the past competitions and have been denied medals because of last minute high scoring kernels. I hope kaggle can take up this concern asap.
\n\nPS. I also have a .9478 stable ensemble submission, but I wont publish it until competition ends. i dont want to ruin genuine kaggler's hardwork by pushing my kernel to public. it kills the interest and the time that lot of genuine kagglers have put through. Same is with other top performing participants.
\n\nAre others feeling the same pain?
", "messages": [ { "id": "893683", "postDate": "06/19/2020 19:55:10", "content": "I just realized my team lost positions in a week (6 positions down today). As a result of posting a high scoring kernel after the deadline, it kinda pisses me off, but i have seen this a lot more times in the past competitions and have been denied medals because of last minute high scoring kernels. I hope kaggle can take up this concern asap.
\n\nPS. I also have a .9478 stable ensemble submission, but I wont publish it until competition ends. i dont want to ruin genuine kaggler's hardwork by pushing my kernel to public. it kills the interest and the time that lot of genuine kagglers have put through. Same is with other top performing participants.
\n\nAre others feeling the same pain?
", "rawMarkdown": "I just realized my team lost positions in a week (6 positions down today). As a result of posting a high scoring kernel after the deadline, it kinda pisses me off, but i have seen this a lot more times in the past competitions and have been denied medals because of last minute high scoring kernels. I hope kaggle can take up this concern asap.\n\nPS. I also have a .9478 stable ensemble submission, but I wont publish it until competition ends. i dont want to ruin genuine kaggler's hardwork by pushing my kernel to public. it kills the interest and the time that lot of genuine kagglers have put through. Same is with other top performing participants. \n\nAre others feeling the same pain?", "votes": null }, { "id": "893812", "postDate": "06/20/2020 00:20:50", "content": "Many people will lose hundreds positions in just a second :D
", "rawMarkdown": "Many people will lose hundreds positions in just a second :D", "votes": null }, { "id": "893887", "postDate": "06/20/2020 04:07:00", "content": "yes absolutly right
", "rawMarkdown": "yes absolutly right", "votes": null }, { "id": "893926", "postDate": "06/20/2020 04:49:10", "content": "agree
", "rawMarkdown": "agree", "votes": null }, { "id": "894029", "postDate": "06/20/2020 05:51:31", "content": "What kernel do you mean? Unfortunately now we should monitoring new high-score kernels to be in the race. If you want to take gold medal then high-score kernels doesn't help much at the end of the competition
", "rawMarkdown": "What kernel do you mean? Unfortunately now we should monitoring new high-score kernels to be in the race. If you want to take gold medal then high-score kernels doesn't help much at the end of the competition", "votes": null }, { "id": "894143", "postDate": "06/20/2020 07:38:20", "content": "i dont get your point, are you favoring a high scoring kernel to be posted in the end days of the competition or you are against it? I am just asking kaggle to administer the high scoring kernels in the last week of the competition, it should be fair game for everyone.
", "rawMarkdown": "i dont get your point, are you favoring a high scoring kernel to be posted in the end days of the competition or you are against it? I am just asking kaggle to administer the high scoring kernels in the last week of the competition, it should be fair game for everyone.", "votes": null }, { "id": "894174", "postDate": "06/20/2020 07:55:50", "content": "if you are seriously working in this comp , knowing how brilliant you are , many people will lose at least one spot that's for sure.
", "rawMarkdown": "if you are seriously working in this comp , knowing how brilliant you are , many people will lose at least one spot that's for sure.", "votes": null }, { "id": "894181", "postDate": "06/20/2020 07:59:48", "content": "same case in trends. but you survived that because your score was a tad better mid silver level kernel that was published by someone in gold zone . I went to low bronze from mid silver in matter of hours . We dont know how much it will matter in private But its not good morally for lot of people. and practically too because 3 days remaining some people want to leverage their position to team up with good people .
", "rawMarkdown": "same case in trends. but you survived that because your score was a tad better mid silver level kernel that was published by someone in gold zone . I went to low bronze from mid silver in matter of hours . We dont know how much it will matter in private But its not good morally for lot of people. and practically too because 3 days remaining some people want to leverage their position to team up with good people .", "votes": null }, { "id": "894234", "postDate": "06/20/2020 08:50:06", "content": "It would be super frustrating if someone's sharing high-score kernels just a few days before the game closes, which definitely kills people when one drops due to that shoot.. I know what you're saying here. But I don't see any high-score notebooks... are they sharing privately?
", "rawMarkdown": "It would be super frustrating if someone's sharing high-score kernels just a few days before the game closes, which definitely kills people when one drops due to that shoot.. I know what you're saying here. But I don't see any high-score notebooks... are they sharing privately?", "votes": null }, { "id": "894245", "postDate": "06/20/2020 09:02:00", "content": "@phoenix9032 thanks for a such a complimentary comment, unfortunately, I've just joined and still can't find whether it better to predict default test or translated to en :)
", "rawMarkdown": "phoenix9032 thanks for a such a complimentary comment, unfortunately, I've just joined and still can't find whether it better to predict default test or translated to en :)", "votes": null }, { "id": "894247", "postDate": "06/20/2020 09:06:04", "content": "I against it. Think we should improve our own models in the last week maybe even two last weeks. Just think that if you really want to take golden medal then such kernels not much helpful. But maybe I am wrong. Will be cool to vote for this point among all kagglers
", "rawMarkdown": "I against it. Think we should improve our own models in the last week maybe even two last weeks. Just think that if you really want to take golden medal then such kernels not much helpful. But maybe I am wrong. Will be cool to vote for this point among all kagglers", "votes": null }, { "id": "894250", "postDate": "06/20/2020 09:07:20", "content": "Predict default and en translation and then average predictions =)
", "rawMarkdown": "Predict default and en translation and then average predictions =)", "votes": null }, { "id": "894258", "postDate": "06/20/2020 09:20:03", "content": "@aybatov shhhhh
", "rawMarkdown": "aybatov shhhhh", "votes": null }, { "id": "894294", "postDate": "06/20/2020 09:59:56", "content": "Which kernel are you speaking about?
", "rawMarkdown": "Which kernel are you speaking about?", "votes": null }, { "id": "894310", "postDate": "06/20/2020 10:26:07", "content": "Same question here. I didn't see any meaningful kernels. Just some weight-tweaking ensembles.
", "rawMarkdown": "Same question here. I didn't see any **meaningful** kernels. Just some weight-tweaking ensembles.", "votes": null }, { "id": "894625", "postDate": "06/20/2020 15:48:27", "content": "The main thing is that, we put in efforts, get a good accuracy, but all of it is invalidated because someone with a high scoring kernel made the same public. I totally agree with OP because this doesnot give rise to healthy competition. Many people will just run/submit those high scoring kernels to obtain a position in the LB.
", "rawMarkdown": "The main thing is that, we put in efforts, get a good accuracy, but all of it is invalidated because someone with a high scoring kernel made the same public. I totally agree with OP because this doesnot give rise to healthy competition. Many people will just run/submit those high scoring kernels to obtain a position in the LB.", "votes": null }, { "id": "895060", "postDate": "06/21/2020 04:42:28", "content": "It's not just because of that, it's because other competitors have started blending their best models.
", "rawMarkdown": "It's not just because of that, it's because other competitors have started blending their best models.", "votes": null }, { "id": "895346", "postDate": "06/21/2020 09:19:19", "content": "So True !!
", "rawMarkdown": "So True !!", "votes": null }, { "id": "900009", "postDate": "06/24/2020 15:14:02", "content": "I gained some 32 odd rank . ya but true other competitors have some of their best models.
", "rawMarkdown": "I gained some 32 odd rank . ya but true other competitors have some of their best models.", "votes": null }, { "id": "900456", "postDate": "06/24/2020 20:07:12", "content": "\n\n\nAs a result of posting a high scoring kernel after the deadline
\n
Which deadline?
", "rawMarkdown": "> As a result of posting a high scoring kernel after the deadline\n\nWhich deadline?", "votes": null }, { "id": "902114", "postDate": "06/26/2020 00:19:16", "content": "The deadline as in when the warning banner comes in - asking kagglers not to post high scoring kernels.
", "rawMarkdown": "The deadline as in when the warning banner comes in - asking kagglers not to post high scoring kernels.", "votes": null } ], "comments": [ { "id": 893812, "author_name": "yaroshevskiy", "author_url": "", "post_date": "06/20/2020 00:20:50", "content": "Many people will lose hundreds positions in just a second :D
", "votes": null, "replies": [ { "id": 894174, "author_name": "phoenix9032", "author_url": "", "post_date": "06/20/2020 07:55:50", "content": "if you are seriously working in this comp , knowing how brilliant you are , many people will lose at least one spot that's for sure.
", "votes": null, "replies": [] }, { "id": 894245, "author_name": "yaroshevskiy", "author_url": "", "post_date": "06/20/2020 09:02:00", "content": "@phoenix9032 thanks for a such a complimentary comment, unfortunately, I've just joined and still can't find whether it better to predict default test or translated to en :)
", "votes": null, "replies": [] }, { "id": 894250, "author_name": "aybatov", "author_url": "", "post_date": "06/20/2020 09:07:20", "content": "Predict default and en translation and then average predictions =)
", "votes": null, "replies": [] }, { "id": 894258, "author_name": "phoenix9032", "author_url": "", "post_date": "06/20/2020 09:20:03", "content": "@aybatov shhhhh
", "votes": null, "replies": [] } ] }, { "id": 893887, "author_name": "muralidhar123", "author_url": "", "post_date": "06/20/2020 04:07:00", "content": "yes absolutly right
", "votes": null, "replies": [] }, { "id": 893926, "author_name": "bavalpreet26", "author_url": "", "post_date": "06/20/2020 04:49:10", "content": "agree
", "votes": null, "replies": [] }, { "id": 894029, "author_name": "aybatov", "author_url": "", "post_date": "06/20/2020 05:51:31", "content": "What kernel do you mean? Unfortunately now we should monitoring new high-score kernels to be in the race. If you want to take gold medal then high-score kernels doesn't help much at the end of the competition
", "votes": null, "replies": [ { "id": 894143, "author_name": "roydatascience", "author_url": "", "post_date": "06/20/2020 07:38:20", "content": "i dont get your point, are you favoring a high scoring kernel to be posted in the end days of the competition or you are against it? I am just asking kaggle to administer the high scoring kernels in the last week of the competition, it should be fair game for everyone.
", "votes": null, "replies": [] }, { "id": 894247, "author_name": "aybatov", "author_url": "", "post_date": "06/20/2020 09:06:04", "content": "I against it. Think we should improve our own models in the last week maybe even two last weeks. Just think that if you really want to take golden medal then such kernels not much helpful. But maybe I am wrong. Will be cool to vote for this point among all kagglers
", "votes": null, "replies": [] }, { "id": 894625, "author_name": "jesudasdsouza", "author_url": "", "post_date": "06/20/2020 15:48:27", "content": "The main thing is that, we put in efforts, get a good accuracy, but all of it is invalidated because someone with a high scoring kernel made the same public. I totally agree with OP because this doesnot give rise to healthy competition. Many people will just run/submit those high scoring kernels to obtain a position in the LB.
", "votes": null, "replies": [] } ] }, { "id": 894181, "author_name": "phoenix9032", "author_url": "", "post_date": "06/20/2020 07:59:48", "content": "same case in trends. but you survived that because your score was a tad better mid silver level kernel that was published by someone in gold zone . I went to low bronze from mid silver in matter of hours . We dont know how much it will matter in private But its not good morally for lot of people. and practically too because 3 days remaining some people want to leverage their position to team up with good people .
", "votes": null, "replies": [] }, { "id": 894234, "author_name": "", "author_url": "", "post_date": "06/20/2020 08:50:06", "content": "It would be super frustrating if someone's sharing high-score kernels just a few days before the game closes, which definitely kills people when one drops due to that shoot.. I know what you're saying here. But I don't see any high-score notebooks... are they sharing privately?
", "votes": null, "replies": [] }, { "id": 894294, "author_name": "philippsinger", "author_url": "", "post_date": "06/20/2020 09:59:56", "content": "Which kernel are you speaking about?
", "votes": null, "replies": [ { "id": 894310, "author_name": "veryrobustperson", "author_url": "", "post_date": "06/20/2020 10:26:07", "content": "Same question here. I didn't see any meaningful kernels. Just some weight-tweaking ensembles.
", "votes": null, "replies": [] } ] }, { "id": 895060, "author_name": "shahules", "author_url": "", "post_date": "06/21/2020 04:42:28", "content": "It's not just because of that, it's because other competitors have started blending their best models.
", "votes": null, "replies": [] }, { "id": 895346, "author_name": "namanj27", "author_url": "", "post_date": "06/21/2020 09:19:19", "content": "So True !!
", "votes": null, "replies": [] }, { "id": 900009, "author_name": "susant4learning", "author_url": "", "post_date": "06/24/2020 15:14:02", "content": "I gained some 32 odd rank . ya but true other competitors have some of their best models.
", "votes": null, "replies": [] }, { "id": 900456, "author_name": "cpmpml", "author_url": "", "post_date": "06/24/2020 20:07:12", "content": "\n\n\nAs a result of posting a high scoring kernel after the deadline
\n
Which deadline?
", "votes": null, "replies": [ { "id": 902114, "author_name": "roydatascience", "author_url": "", "post_date": "06/26/2020 00:19:16", "content": "The deadline as in when the warning banner comes in - asking kagglers not to post high scoring kernels.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "893683": "I just realized my team lost positions in a week (6 positions down today). As a result of posting a high scoring kernel after the deadline, it kinda pisses me off, but i have seen this a lot more times in the past competitions and have been denied medals because of last minute high scoring kernels. I hope kaggle can take up this concern asap.\n\nPS. I also have a .9478 stable ensemble submission, but I wont publish it until competition ends. i dont want to ruin genuine kaggler's hardwork by pushing my kernel to public. it kills the interest and the time that lot of genuine kagglers have put through. Same is with other top performing participants. \n\nAre others feeling the same pain?", "893812": "Many people will lose hundreds positions in just a second :D", "893887": "yes absolutly right", "893926": "agree", "894029": "What kernel do you mean? Unfortunately now we should monitoring new high-score kernels to be in the race. If you want to take gold medal then high-score kernels doesn't help much at the end of the competition", "894143": "i dont get your point, are you favoring a high scoring kernel to be posted in the end days of the competition or you are against it? I am just asking kaggle to administer the high scoring kernels in the last week of the competition, it should be fair game for everyone.", "894174": "if you are seriously working in this comp , knowing how brilliant you are , many people will lose at least one spot that's for sure.", "894181": "same case in trends. but you survived that because your score was a tad better mid silver level kernel that was published by someone in gold zone . I went to low bronze from mid silver in matter of hours . We dont know how much it will matter in private But its not good morally for lot of people. and practically too because 3 days remaining some people want to leverage their position to team up with good people .", "894234": "It would be super frustrating if someone's sharing high-score kernels just a few days before the game closes, which definitely kills people when one drops due to that shoot.. I know what you're saying here. But I don't see any high-score notebooks... are they sharing privately?", "894245": "phoenix9032 thanks for a such a complimentary comment, unfortunately, I've just joined and still can't find whether it better to predict default test or translated to en :)", "894247": "I against it. Think we should improve our own models in the last week maybe even two last weeks. Just think that if you really want to take golden medal then such kernels not much helpful. But maybe I am wrong. Will be cool to vote for this point among all kagglers", "894250": "Predict default and en translation and then average predictions =)", "894258": "aybatov shhhhh", "894294": "Which kernel are you speaking about?", "894310": "Same question here. I didn't see any **meaningful** kernels. Just some weight-tweaking ensembles.", "894625": "The main thing is that, we put in efforts, get a good accuracy, but all of it is invalidated because someone with a high scoring kernel made the same public. I totally agree with OP because this doesnot give rise to healthy competition. Many people will just run/submit those high scoring kernels to obtain a position in the LB.", "895060": "It's not just because of that, it's because other competitors have started blending their best models.", "895346": "So True !!", "900009": "I gained some 32 odd rank . ya but true other competitors have some of their best models.", "900456": "> As a result of posting a high scoring kernel after the deadline\n\nWhich deadline?", "902114": "The deadline as in when the warning banner comes in - asking kagglers not to post high scoring kernels." }, "source": "meta" }