I am all for sharing of notebooks, but over 20% of the competition appear to have just submitted @corochann 's kernel - this doesn't seem to be in the spirit of Kaggle?
", "messages": [ { "id": 728969, "postDate": "2020-01-25T14:14:15.727Z", "content": "I am all for sharing of notebooks, but over 20% of the competition appear to have just submitted @corochann 's kernel - this doesn't seem to be in the spirit of Kaggle?
", "rawMarkdown": "I am all for sharing of notebooks, but over 20% of the competition appear to have just submitted @corochann 's kernel - this doesn't seem to be in the spirit of Kaggle?", "votes": 22 }, { "id": 729577, "postDate": "2020-01-26T11:56:05.630Z", "content": "Don't be frustrated, 2 months to go, they will end up way lower.\nKeep on trying and learning and you will pass them.
", "rawMarkdown": "Don't be frustrated, 2 months to go, they will end up way lower.\nKeep on trying and learning and you will pass them.", "votes": 5 }, { "id": 729626, "postDate": "2020-01-26T13:19:18.183Z", "content": "> this doesn't seem to be in the spirit of Kaggle?
\n\nSorry, but this is the spirit of Kaggle for most of the participants, sadly. In every competition hundreds of people fork and submit top-scoring kernels.
", "rawMarkdown": "> this doesn't seem to be in the spirit of Kaggle?\n\nSorry, but this is the spirit of Kaggle for most of the participants, sadly. In every competition hundreds of people fork and submit top-scoring kernels.", "votes": 6, "replies": [ { "id": 740984, "postDate": "2020-02-10T03:39:57.153Z", "content": "but they rarely get medals, and there are some others who stacks top-scoring kernels and gets medals.
", "rawMarkdown": "but they rarely get medals, and there are some others who stacks top-scoring kernels and gets medals.\n" } ] }, { "id": 732004, "postDate": "2020-01-29T11:22:07.427Z", "content": "A public kernel can never win a medal. Don't worry about this. Just consider it as a baseline and try to surpass it or build your own ideas on it.
", "rawMarkdown": "A public kernel can never win a medal. Don't worry about this. Just consider it as a baseline and try to surpass it or build your own ideas on it.", "votes": 4, "replies": [ { "id": 741426, "postDate": "2020-02-10T16:09:15.783Z", "content": "\n\n\nA public kernel can never win a medal.
\n
They do sometimes (e.g. instant gratification competition)
", "rawMarkdown": "> A public kernel can never win a medal.\n\nThey do sometimes (e.g. instant gratification competition)", "votes": 1 } ] }, { "id": 730032, "postDate": "2020-01-27T02:38:50.300Z", "content": "dont worry. nobody get a medal just by re-submit public kernels
", "rawMarkdown": "dont worry. nobody get a medal just by re-submit public kernels", "votes": 3 }, { "id": 744523, "postDate": "2020-02-12T22:44:52.867Z", "content": "175 people now copied .9696 and there's still 200+ at .9663 lol. If I increase my score .002 I would go up 200 spots!\nEdit: Look ma! I did it!
", "rawMarkdown": "175 people now copied .9696 and there's still 200+ at .9663 lol. If I increase my score .002 I would go up 200 spots!\nEdit: Look ma! I did it!", "votes": 1 }, { "id": 730276, "postDate": "2020-01-27T09:55:59.487Z", "content": "I have mixed feelings on this. I think being able to share high-scoring kernels is important for advancing knowledge especially for people to find short-comings they may have missed. Ultimately, it's up to Kaggle to decide how to do this - perhaps kernels which are currently medal winning would be exempt from being shared?
\n\nOtherwise, Kaggle shouldn't be able to say \"oh, your kernel is too good to share\" etc. because that's holding back knowledge that could certainly help others learn/improve.
", "rawMarkdown": "I have mixed feelings on this. I think being able to share high-scoring kernels is important for advancing knowledge especially for people to find short-comings they may have missed. Ultimately, it's up to Kaggle to decide how to do this - perhaps kernels which are currently medal winning would be exempt from being shared? \n\nOtherwise, Kaggle shouldn't be able to say \"oh, your kernel is too good to share\" etc. because that's holding back knowledge that could certainly help others learn/improve.", "votes": 1 }, { "id": 736305, "postDate": "2020-02-04T03:45:46.973Z", "content": "People share kernels to share knowledge. It should not be copied as it is. Atleast put it in the effort of understanding the technique and try improving over it instead of simply copying.\nP.S my latest score is 0.9643 😟
", "rawMarkdown": "People share kernels to share knowledge. It should not be copied as it is. Atleast put it in the effort of understanding the technique and try improving over it instead of simply copying.\nP.S my latest score is 0.9643 😟 ", "votes": 2 }, { "id": 729790, "postDate": "2020-01-26T16:39:36.413Z", "content": "I think the intention of sharing kernel is to share knowledge.
\n\nI have seen people participate in the competition and without even reading the problem statement fork the high scoring kernel, submit it and gets dormant.
", "rawMarkdown": "I think the intention of sharing kernel is to share knowledge.\n\nI have seen people participate in the competition and without even reading the problem statement fork the high scoring kernel, submit it and gets dormant. ", "votes": 2 }, { "id": 729759, "postDate": "2020-01-26T16:11:17.220Z", "content": "Agreed. It's one thing to learn from the work of others. It's another to submit it as your own and benefit from it.
", "rawMarkdown": "Agreed. It's one thing to learn from the work of others. It's another to submit it as your own and benefit from it.", "votes": 2 }, { "id": 729482, "postDate": "2020-01-26T09:02:07.667Z", "content": "Completely agreed on this. This even gets you in bronze.
\n\nHopefully, there's going to be a bit of shake-up on the private LB, and all of these are going to go either up or down. I'm guessing having a good cross-validation is a good way to pass above those 20% of the contestants.
", "rawMarkdown": "Completely agreed on this. This even gets you in bronze.\n\nHopefully, there's going to be a bit of shake-up on the private LB, and all of these are going to go either up or down. I'm guessing having a good cross-validation is a good way to pass above those 20% of the contestants.", "votes": 2 }, { "id": 731113, "postDate": "2020-01-28T11:09:20.913Z", "content": "I understand your frustration, but how is it against Kaggle spirit? Kernels and public discussions are great tools and provide opportunity to learn for many people, and just as with any other tool, it's totally up to you how to use it. And if someone just wants to submit without any understanding whatsoever, so be it. Anyway, as @moewie94 pointed out, nobody get a medal just by re-submitting public kernels
", "rawMarkdown": "I understand your frustration, but how is it against Kaggle spirit? Kernels and public discussions are great tools and provide opportunity to learn for many people, and just as with any other tool, it's totally up to you how to use it. And if someone just wants to submit without any understanding whatsoever, so be it. Anyway, as @moewie94 pointed out, nobody get a medal just by re-submitting public kernels" }, { "id": 734526, "postDate": "2020-02-01T16:06:37.390Z", "rawMarkdown": "", "isDeleted": true }, { "id": 729585, "postDate": "2020-01-26T12:08:27.843Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 729577, "author_name": "Max", "author_url": "", "post_date": "2020-01-26T11:56:05.630000", "content": "Don't be frustrated, 2 months to go, they will end up way lower.\nKeep on trying and learning and you will pass them.
", "votes": 5, "replies": [] }, { "id": 729626, "author_name": "Andrey Lukyanenko", "author_url": "", "post_date": "2020-01-26T13:19:18.183000", "content": "> this doesn't seem to be in the spirit of Kaggle?
\n\nSorry, but this is the spirit of Kaggle for most of the participants, sadly. In every competition hundreds of people fork and submit top-scoring kernels.
", "votes": 6, "replies": [ { "id": 740984, "author_name": "Shahules", "author_url": "", "post_date": "2020-02-10T03:39:57.153000", "content": "but they rarely get medals, and there are some others who stacks top-scoring kernels and gets medals.
", "votes": 0, "replies": [] } ] }, { "id": 732004, "author_name": "syoya", "author_url": "", "post_date": "2020-01-29T11:22:07.427000", "content": "A public kernel can never win a medal. Don't worry about this. Just consider it as a baseline and try to surpass it or build your own ideas on it.
", "votes": 4, "replies": [ { "id": 741426, "author_name": "Marco Gorelli", "author_url": "", "post_date": "2020-02-10T16:09:15.783000", "content": "\n\n\nA public kernel can never win a medal.
\n
They do sometimes (e.g. instant gratification competition)
", "votes": 1, "replies": [] } ] }, { "id": 730032, "author_name": "DatNT", "author_url": "", "post_date": "2020-01-27T02:38:50.300000", "content": "dont worry. nobody get a medal just by re-submit public kernels
", "votes": 3, "replies": [] }, { "id": 744523, "author_name": "GreatGameDota", "author_url": "", "post_date": "2020-02-12T22:44:52.867000", "content": "175 people now copied .9696 and there's still 200+ at .9663 lol. If I increase my score .002 I would go up 200 spots!\nEdit: Look ma! I did it!
", "votes": 1, "replies": [] }, { "id": 730276, "author_name": "Nina Sundown", "author_url": "", "post_date": "2020-01-27T09:55:59.487000", "content": "I have mixed feelings on this. I think being able to share high-scoring kernels is important for advancing knowledge especially for people to find short-comings they may have missed. Ultimately, it's up to Kaggle to decide how to do this - perhaps kernels which are currently medal winning would be exempt from being shared?
\n\nOtherwise, Kaggle shouldn't be able to say \"oh, your kernel is too good to share\" etc. because that's holding back knowledge that could certainly help others learn/improve.
", "votes": 1, "replies": [] }, { "id": 736305, "author_name": "Pawan Chorasiya", "author_url": "", "post_date": "2020-02-04T03:45:46.973000", "content": "People share kernels to share knowledge. It should not be copied as it is. Atleast put it in the effort of understanding the technique and try improving over it instead of simply copying.\nP.S my latest score is 0.9643 😟
", "votes": 2, "replies": [] }, { "id": 729790, "author_name": "Amit", "author_url": "", "post_date": "2020-01-26T16:39:36.413000", "content": "I think the intention of sharing kernel is to share knowledge.
\n\nI have seen people participate in the competition and without even reading the problem statement fork the high scoring kernel, submit it and gets dormant.
", "votes": 2, "replies": [] }, { "id": 729759, "author_name": "Mike Schubert", "author_url": "", "post_date": "2020-01-26T16:11:17.220000", "content": "Agreed. It's one thing to learn from the work of others. It's another to submit it as your own and benefit from it.
", "votes": 2, "replies": [] }, { "id": 729482, "author_name": "Maxime Lenormand", "author_url": "", "post_date": "2020-01-26T09:02:07.667000", "content": "Completely agreed on this. This even gets you in bronze.
\n\nHopefully, there's going to be a bit of shake-up on the private LB, and all of these are going to go either up or down. I'm guessing having a good cross-validation is a good way to pass above those 20% of the contestants.
", "votes": 2, "replies": [] }, { "id": 731113, "author_name": "Eek The Cat", "author_url": "", "post_date": "2020-01-28T11:09:20.913000", "content": "I understand your frustration, but how is it against Kaggle spirit? Kernels and public discussions are great tools and provide opportunity to learn for many people, and just as with any other tool, it's totally up to you how to use it. And if someone just wants to submit without any understanding whatsoever, so be it. Anyway, as @moewie94 pointed out, nobody get a medal just by re-submitting public kernels
", "votes": 0, "replies": [] }, { "id": 734526, "author_name": "", "author_url": "", "post_date": "2020-02-01T16:06:37.390000", "content": "", "votes": 0, "replies": [] }, { "id": 729585, "author_name": "", "author_url": "", "post_date": "2020-01-26T12:08:27.843000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "728969": "I am all for sharing of notebooks, but over 20% of the competition appear to have just submitted @corochann 's kernel - this doesn't seem to be in the spirit of Kaggle?", "729577": "Don't be frustrated, 2 months to go, they will end up way lower.\nKeep on trying and learning and you will pass them.", "729626": "> this doesn't seem to be in the spirit of Kaggle?\n\nSorry, but this is the spirit of Kaggle for most of the participants, sadly. In every competition hundreds of people fork and submit top-scoring kernels.", "732004": "A public kernel can never win a medal. Don't worry about this. Just consider it as a baseline and try to surpass it or build your own ideas on it.", "730032": "dont worry. nobody get a medal just by re-submit public kernels", "744523": "175 people now copied .9696 and there's still 200+ at .9663 lol. If I increase my score .002 I would go up 200 spots!\nEdit: Look ma! I did it!", "730276": "I have mixed feelings on this. I think being able to share high-scoring kernels is important for advancing knowledge especially for people to find short-comings they may have missed. Ultimately, it's up to Kaggle to decide how to do this - perhaps kernels which are currently medal winning would be exempt from being shared? \n\nOtherwise, Kaggle shouldn't be able to say \"oh, your kernel is too good to share\" etc. because that's holding back knowledge that could certainly help others learn/improve.", "736305": "People share kernels to share knowledge. It should not be copied as it is. Atleast put it in the effort of understanding the technique and try improving over it instead of simply copying.\nP.S my latest score is 0.9643 😟 ", "729790": "I think the intention of sharing kernel is to share knowledge.\n\nI have seen people participate in the competition and without even reading the problem statement fork the high scoring kernel, submit it and gets dormant. ", "729759": "Agreed. It's one thing to learn from the work of others. It's another to submit it as your own and benefit from it.", "729482": "Completely agreed on this. This even gets you in bronze.\n\nHopefully, there's going to be a bit of shake-up on the private LB, and all of these are going to go either up or down. I'm guessing having a good cross-validation is a good way to pass above those 20% of the contestants.", "731113": "I understand your frustration, but how is it against Kaggle spirit? Kernels and public discussions are great tools and provide opportunity to learn for many people, and just as with any other tool, it's totally up to you how to use it. And if someone just wants to submit without any understanding whatsoever, so be it. Anyway, as @moewie94 pointed out, nobody get a medal just by re-submitting public kernels", "734526": "", "729585": "" } }