Disclaimer: I have not entered this competition, not even looked at data.
\n\nBy quickly going through the first few threads in the Discussion, it seems that this competition consists of:
\n\nIs this the most toxic kaggle competition ever?
", "messages": [ { "id": 699029, "postDate": "2019-12-20T01:36:29.537Z", "content": "Disclaimer: I have not entered this competition, not even looked at data.
\n\nBy quickly going through the first few threads in the Discussion, it seems that this competition consists of:
\n\nIs this the most toxic kaggle competition ever?
", "rawMarkdown": "Disclaimer: I have not entered this competition, not even looked at data.\n\nBy quickly going through the first few threads in the Discussion, it seems that this competition consists of:\n\n- Massive leakage.\n- Suspicious private sharing.\n- Suspicious fake accounts (to probe LB?).\n\nIs this the most toxic kaggle competition ever? ", "votes": 25 }, { "id": 699036, "postDate": "2019-12-20T01:49:40.190Z", "content": "Almost 4000 teams decrease to 1800 and the number is still getting smaller.
", "rawMarkdown": "Almost 4000 teams decrease to 1800 and the number is still getting smaller.", "votes": 5, "replies": [ { "id": 699076, "postDate": "2019-12-20T03:00:58.383Z", "content": "Yeah, that is kinda strange, from around 3700 to 1877 teams, but it is still showing all the teams on leaderboard.
", "rawMarkdown": "Yeah, that is kinda strange, from around 3700 to 1877 teams, but it is still showing all the teams on leaderboard.", "votes": 2 }, { "id": 699210, "postDate": "2019-12-20T07:03:09.313Z", "content": "Are they removed? If so this is insane. I wonder what is special with this comp that attracts 1000s of fake accounts.
", "rawMarkdown": "Are they removed? If so this is insane. I wonder what is special with this comp that attracts 1000s of fake accounts.", "votes": 1, "isDeleted": true }, { "id": 699246, "postDate": "2019-12-20T08:30:28.467Z", "content": "Max 2 subs per day. People may create fake accounts to get more submissions.
", "rawMarkdown": "Max 2 subs per day. People may create fake accounts to get more submissions. " }, { "id": 699247, "postDate": "2019-12-20T08:32:06.810Z", "content": "I think it's just an artefact of rerunning all the submissions on cleared data, I hope it returns back to normal in a few hours/days
", "rawMarkdown": "I think it's just an artefact of rerunning all the submissions on cleared data, I hope it returns back to normal in a few hours/days" }, { "id": 699610, "postDate": "2019-12-20T16:26:23.173Z", "content": "Yes only 2 subs a day and 2 decimal digits for precision.
\n\nIt was a very strange competition, but 4000 teams and now 1800, itβs incredible !
\n\nSo many cheaters ? Hum... If yes, there is something to do tracking them from the beginning of competition.
\n\nI know this is not very easy...
", "rawMarkdown": "Yes only 2 subs a day and 2 decimal digits for precision.\n\nIt was a very strange competition, but 4000 teams and now 1800, itβs incredible ! \n\nSo many cheaters ? Hum... If yes, there is something to do tracking them from the beginning of competition.\n\nI know this is not very easy..." } ] }, { "id": 699225, "postDate": "2019-12-20T07:51:05.120Z", "content": "It was my first competition so honestly I did not know what to expect. Seeing the number of teams almost cut by half is sad. I learned many things from the amazing kernels that were shared in this competiton. \nTo all amazing kagglers, please keep sharing kernels and bring interesting insights after the end of this competiton. Thank you all !
", "rawMarkdown": "It was my first competition so honestly I did not know what to expect. Seeing the number of teams almost cut by half is sad. I learned many things from the amazing kernels that were shared in this competiton. \nTo all amazing kagglers, please keep sharing kernels and bring interesting insights after the end of this competiton. Thank you all !", "votes": 3, "replies": [ { "id": 699668, "postDate": "2019-12-20T18:19:38.360Z", "content": "Same here!! Congrats on your learning progress! Thanks everyone for the open notebooks have taught me so much.
", "rawMarkdown": "Same here!! Congrats on your learning progress! Thanks everyone for the open notebooks have taught me so much. " }, { "id": 699679, "postDate": "2019-12-20T18:48:11.827Z", "content": "Congrats, you got the most valuable aspect of Kaggle IMHO: learning something new!
", "rawMarkdown": "Congrats, you got the most valuable aspect of Kaggle IMHO: learning something new! " } ] }, { "id": 699046, "postDate": "2019-12-20T02:16:10.857Z", "content": "Honestly it's sad. It had so much potential. The data was really cool, especially because one had to learn a lot about Outlier analysis, time series and how to deal with a lot of data. It was my first competition so I learned a lot nonetheless, but I see how it could've gone so much better...
", "rawMarkdown": "Honestly it's sad. It had so much potential. The data was really cool, especially because one had to learn a lot about Outlier analysis, time series and how to deal with a lot of data. It was my first competition so I learned a lot nonetheless, but I see how it could've gone so much better...", "votes": 4, "replies": [ { "id": 699053, "postDate": "2019-12-20T02:21:46.667Z", "content": "It seems you've learn a lot from this competition. That's great. I didn't even follow what you have done. You have many potentials. keep going! ;-)
", "rawMarkdown": "It seems you've learn a lot from this competition. That's great. I didn't even follow what you have done. You have many potentials. keep going! ;-)", "votes": 1 }, { "id": 699060, "postDate": "2019-12-20T02:32:51.240Z", "content": "Thanks for the nice words dude :`), that's really nice to hear. I did learn tons! Mb we can team up in future competitions!
", "rawMarkdown": "Thanks for the nice words dude :`), that's really nice to hear. I did learn tons! Mb we can team up in future competitions!", "votes": 1 }, { "id": 699064, "postDate": "2019-12-20T02:41:49.243Z", "content": "No problem, Someday both of us could get strong backgrounds enough to get a synergy ;-)
", "rawMarkdown": "No problem, Someday both of us could get strong backgrounds enough to get a synergy ;-)", "votes": 1 } ] }, { "id": 699102, "postDate": "2019-12-20T03:45:18.423Z", "content": "I respectfully disagree. Here's a thread that you probably missed: https://www.kaggle.com/c/ashrae-energy-prediction/discussion/119460
\n\nDespite the problems, there will be no damage in the private leaderboard and measures are being taken regarding suspicious or forbidden activities.
", "rawMarkdown": "I respectfully disagree. Here's a thread that you probably missed: https://www.kaggle.com/c/ashrae-energy-prediction/discussion/119460\n\nDespite the problems, there will be no damage in the private leaderboard and measures are being taken regarding suspicious or forbidden activities. ", "votes": 2, "replies": [ { "id": 699151, "postDate": "2019-12-20T05:18:33.957Z", "content": "I disagree as well. I actually had fun and learned a lot.
", "rawMarkdown": "I disagree as well. I actually had fun and learned a lot.", "votes": 5 }, { "id": 699248, "postDate": "2019-12-20T08:32:16.320Z", "content": "It was the same for me π
", "rawMarkdown": "It was the same for me π " }, { "id": 699267, "postDate": "2019-12-20T08:50:33.537Z", "content": "Yes, removing leakage from the private leaderboard minimizes the damage, but it can not removing the impact of leakage. For example, people need to spend extra effort to collect the leakage. If you do not hold all the leakages, you may be at a bad position. You may argue that collecting data is one part of data science, which I do agree, but I don't think it should belong to a data science competition.
\n\nAs for removing suspicious of forbidden activities, you could expect some will be taken care of, but do not expect all of them.
", "rawMarkdown": "Yes, removing leakage from the private leaderboard minimizes the damage, but it can not removing the impact of leakage. For example, people need to spend extra effort to collect the leakage. If you do not hold all the leakages, you may be at a bad position. You may argue that collecting data is one part of data science, which I do agree, but I don't think it should belong to a data science competition. \n\nAs for removing suspicious of forbidden activities, you could expect some will be taken care of, but do not expect all of them. " }, { "id": 699271, "postDate": "2019-12-20T08:56:43.740Z", "content": "+1. People like to complain and probably there is a reason to complain in this comp but overall it was quite interesting.
", "rawMarkdown": "+1. People like to complain and probably there is a reason to complain in this comp but overall it was quite interesting.", "votes": 1 }, { "id": 699421, "postDate": "2019-12-20T12:43:37.313Z", "content": "I disagree as well.\nAlthough a lot of leakage data in this game, there are still more than three thousand people participate in this competition.
", "rawMarkdown": "I disagree as well.\nAlthough a lot of leakage data in this game, there are still more than three thousand people participate in this competition." }, { "id": 699558, "postDate": "2019-12-20T15:37:16.230Z", "content": "\n\n\nFor example, people need to spend extra effort to collect the leakage.
\n
@naivelamb since you were not participating, probably you are not aware but some great Kagglers (@gunesevitan, @mpware, @poedator, @pdnartreb, @serengil and @pp2file) thankfully did the hard work of collecting the leakage. Also, the Kaggler @yamsam created a great kernel joining all the leakages. Thanks to them, the impact was minimal.
", "rawMarkdown": "> For example, people need to spend extra effort to collect the leakage.\n\n@naivelamb since you were not participating, probably you are not aware but some great Kagglers (@gunesevitan, @mpware, @poedator, @pdnartreb, @serengil and @pp2file) thankfully did the hard work of collecting the leakage. Also, the Kaggler @yamsam created a [great kernel](https://www.kaggle.com/yamsam/ashrae-leak-data-station) joining all the leakages. Thanks to them, the impact was minimal. ", "votes": 3 }, { "id": 699690, "postDate": "2019-12-20T19:10:36.427Z", "content": "I agree for all the competitions with leakages, there are great kagglers helping to exploit the leakages and publish it. Shout out to them. However, the question still hangs there: are all the leakages floating on the Internet being published?
\n\nI remembered that in the Google Analytics competition, when the Google Demo Account can be used to mining the target variables, lots of people left the competition and Kaggle had to reformulate it. In this competition, lots of people also left after releasing the target variables can be mined.
\n\nI think the impact of leakage can be estimate by asking one question:\n- Will you join a competition knowing there are leakages?
\n\nIf majority people answer 'No', its impact cannot be ignored.
\n\nBTW, congrats on you 18th position on public leaderboard, wish you get a good one on the private leaderboard as well.
", "rawMarkdown": "I agree for all the competitions with leakages, there are great kagglers helping to exploit the leakages and publish it. Shout out to them. However, the question still hangs there: are all the leakages floating on the Internet being published? \n\nI remembered that in the [Google Analytics competition](https://www.kaggle.com/c/ga-customer-revenue-prediction), when the Google Demo Account can be used to mining the target variables, lots of people left the competition and Kaggle had to reformulate it. In this competition, lots of people also left after releasing the target variables can be mined. \n\nI think the impact of leakage can be estimate by asking one question:\n- Will you join a competition knowing there are leakages?\n\nIf majority people answer 'No', its impact cannot be ignored. \n\nBTW, congrats on you 18th position on public leaderboard, wish you get a good one on the private leaderboard as well. " } ] }, { "id": 717373, "postDate": "2020-01-13T04:05:01.313Z", "content": "Hey all! On behalf of the technical committee from ASHRAE, I just wanted to thank you all for your feedback and for sticking in there through the competition. We hope you all learned something despite the issues raised. A lot of cool analysis was created and shared through the process and we're happy for that. We learned a lot about how to host and the next one will be better, don't worry :)
", "rawMarkdown": "Hey all! On behalf of the technical committee from ASHRAE, I just wanted to thank you all for your feedback and for sticking in there through the competition. We hope you all learned something despite the issues raised. A lot of cool analysis was created and shared through the process and we're happy for that. We learned a lot about how to host and the next one will be better, don't worry :)", "votes": 1 }, { "id": 699164, "postDate": "2019-12-20T05:41:50.737Z", "content": "I believe we have seen competitions where a high scoring kernel was posted at the end or people were accused of private sharing or fake accounts (which can be true too). People will do what they wanna do, kaggle can only put regulations and do compliance to get rid of above.
\n\nThis competition differs because of leak data, but kaggle is trying to get rid of that so that it wonβt affect the final standings.
\n\nOn lighter note, I still believe that Jigsaw Toxic classification competition holds the first spot π
", "rawMarkdown": "I believe we have seen competitions where a high scoring kernel was posted at the end or people were accused of private sharing or fake accounts (which can be true too). People will do what they wanna do, kaggle can only put regulations and do compliance to get rid of above. \n\nThis competition differs because of leak data, but kaggle is trying to get rid of that so that it wonβt affect the final standings.\n\nOn lighter note, I still believe that Jigsaw Toxic classification competition holds the first spot π", "votes": 1 }, { "id": 699561, "postDate": "2019-12-20T15:38:43.627Z", "content": "The worst competition ever may be this one : https://www.kaggle.com/c/draper-satellite-image-chronology/overview.
\n\nAll the gold medals, a priori, got it by manually sorting the pictures without any ML or any calculation at all.
", "rawMarkdown": "The worst competition ever may be this one : https://www.kaggle.com/c/draper-satellite-image-chronology/overview.\n\nAll the gold medals, a priori, got it by manually sorting the pictures without any ML or any calculation at all.", "votes": 2, "replies": [ { "id": 699756, "postDate": "2019-12-20T21:23:13.053Z", "content": "Luckily I wasn't playing kaggle at that time. π
", "rawMarkdown": "Luckily I wasn't playing kaggle at that time. π " } ] }, { "id": 699734, "postDate": "2019-12-20T20:35:22.703Z", "content": "These issues were predictable, see for instance: https://www.kaggle.com/c/ashrae-energy-prediction/discussion/116840
", "rawMarkdown": "These issues were predictable, see for instance: https://www.kaggle.com/c/ashrae-energy-prediction/discussion/116840", "replies": [ { "id": 699755, "postDate": "2019-12-20T21:22:28.337Z", "content": "Yeah, I saw your post and decided not to join this competition.
", "rawMarkdown": "Yeah, I saw your post and decided not to join this competition. ", "votes": 1 } ] }, { "id": 699236, "postDate": "2019-12-20T08:15:15.180Z", "content": "'Last minute high scoring kernel.'
\n\nCan I ask where is it?
", "rawMarkdown": "'Last minute high scoring kernel.'\n\nCan I ask where is it?", "replies": [ { "id": 699258, "postDate": "2019-12-20T08:42:37.820Z", "content": "It was mentioned in some of the threads I went through. Unfortunately, it seems to be deleted.
", "rawMarkdown": "It was mentioned in some of the threads I went through. Unfortunately, it seems to be deleted. " }, { "id": 699293, "postDate": "2019-12-20T09:32:03.890Z", "content": "@naivelamb I would request you to go through this discussion and kindly edit yours. :)\nhttps://www.kaggle.com/c/ashrae-energy-prediction/discussion/122446\nThank you.
", "rawMarkdown": "@naivelamb I would request you to go through this discussion and kindly edit yours. :)\nhttps://www.kaggle.com/c/ashrae-energy-prediction/discussion/122446\nThank you. " }, { "id": 699601, "postDate": "2019-12-20T16:18:06.590Z", "content": "OK, edited.
", "rawMarkdown": "OK, edited. ", "votes": 1 } ] }, { "id": 699259, "postDate": "2019-12-20T08:44:04.430Z", "rawMarkdown": "", "isDeleted": true }, { "id": 699047, "postDate": "2019-12-20T02:16:51.057Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 699036, "author_name": "fwj7", "author_url": "", "post_date": "2019-12-20T01:49:40.190000", "content": "Almost 4000 teams decrease to 1800 and the number is still getting smaller.
", "votes": 5, "replies": [ { "id": 699076, "author_name": "HarshitMehta", "author_url": "", "post_date": "2019-12-20T03:00:58.383000", "content": "Yeah, that is kinda strange, from around 3700 to 1877 teams, but it is still showing all the teams on leaderboard.
", "votes": 2, "replies": [] }, { "id": 699210, "author_name": "", "author_url": "", "post_date": "2019-12-20T07:03:09.313000", "content": "Are they removed? If so this is insane. I wonder what is special with this comp that attracts 1000s of fake accounts.
", "votes": 1, "replies": [] }, { "id": 699246, "author_name": "Xuan Cao", "author_url": "", "post_date": "2019-12-20T08:30:28.467000", "content": "Max 2 subs per day. People may create fake accounts to get more submissions.
", "votes": 0, "replies": [] }, { "id": 699247, "author_name": "Fred Navruzov", "author_url": "", "post_date": "2019-12-20T08:32:06.810000", "content": "I think it's just an artefact of rerunning all the submissions on cleared data, I hope it returns back to normal in a few hours/days
", "votes": 0, "replies": [] }, { "id": 699610, "author_name": "mezoganet", "author_url": "", "post_date": "2019-12-20T16:26:23.173000", "content": "Yes only 2 subs a day and 2 decimal digits for precision.
\n\nIt was a very strange competition, but 4000 teams and now 1800, itβs incredible !
\n\nSo many cheaters ? Hum... If yes, there is something to do tracking them from the beginning of competition.
\n\nI know this is not very easy...
", "votes": 0, "replies": [] } ] }, { "id": 699225, "author_name": "Nono", "author_url": "", "post_date": "2019-12-20T07:51:05.120000", "content": "It was my first competition so honestly I did not know what to expect. Seeing the number of teams almost cut by half is sad. I learned many things from the amazing kernels that were shared in this competiton. \nTo all amazing kagglers, please keep sharing kernels and bring interesting insights after the end of this competiton. Thank you all !
", "votes": 3, "replies": [ { "id": 699668, "author_name": "DavidHam", "author_url": "", "post_date": "2019-12-20T18:19:38.360000", "content": "Same here!! Congrats on your learning progress! Thanks everyone for the open notebooks have taught me so much.
", "votes": 0, "replies": [] }, { "id": 699679, "author_name": "Xuan Cao", "author_url": "", "post_date": "2019-12-20T18:48:11.827000", "content": "Congrats, you got the most valuable aspect of Kaggle IMHO: learning something new!
", "votes": 0, "replies": [] } ] }, { "id": 699046, "author_name": "Thiago Preischadt", "author_url": "", "post_date": "2019-12-20T02:16:10.857000", "content": "Honestly it's sad. It had so much potential. The data was really cool, especially because one had to learn a lot about Outlier analysis, time series and how to deal with a lot of data. It was my first competition so I learned a lot nonetheless, but I see how it could've gone so much better...
", "votes": 4, "replies": [ { "id": 699053, "author_name": "Hanjoon Choe", "author_url": "", "post_date": "2019-12-20T02:21:46.667000", "content": "It seems you've learn a lot from this competition. That's great. I didn't even follow what you have done. You have many potentials. keep going! ;-)
", "votes": 1, "replies": [] }, { "id": 699060, "author_name": "Thiago Preischadt", "author_url": "", "post_date": "2019-12-20T02:32:51.240000", "content": "Thanks for the nice words dude :`), that's really nice to hear. I did learn tons! Mb we can team up in future competitions!
", "votes": 1, "replies": [] }, { "id": 699064, "author_name": "Hanjoon Choe", "author_url": "", "post_date": "2019-12-20T02:41:49.243000", "content": "No problem, Someday both of us could get strong backgrounds enough to get a synergy ;-)
", "votes": 1, "replies": [] } ] }, { "id": 699102, "author_name": "Fernando Wittmann", "author_url": "", "post_date": "2019-12-20T03:45:18.423000", "content": "I respectfully disagree. Here's a thread that you probably missed: https://www.kaggle.com/c/ashrae-energy-prediction/discussion/119460
\n\nDespite the problems, there will be no damage in the private leaderboard and measures are being taken regarding suspicious or forbidden activities.
", "votes": 2, "replies": [ { "id": 699151, "author_name": "Gunes Evitan", "author_url": "", "post_date": "2019-12-20T05:18:33.957000", "content": "I disagree as well. I actually had fun and learned a lot.
", "votes": 5, "replies": [] }, { "id": 699248, "author_name": "Mimou", "author_url": "", "post_date": "2019-12-20T08:32:16.320000", "content": "It was the same for me π
", "votes": 0, "replies": [] }, { "id": 699267, "author_name": "Xuan Cao", "author_url": "", "post_date": "2019-12-20T08:50:33.537000", "content": "Yes, removing leakage from the private leaderboard minimizes the damage, but it can not removing the impact of leakage. For example, people need to spend extra effort to collect the leakage. If you do not hold all the leakages, you may be at a bad position. You may argue that collecting data is one part of data science, which I do agree, but I don't think it should belong to a data science competition.
\n\nAs for removing suspicious of forbidden activities, you could expect some will be taken care of, but do not expect all of them.
", "votes": 0, "replies": [] }, { "id": 699271, "author_name": "Artem Dubinich", "author_url": "", "post_date": "2019-12-20T08:56:43.740000", "content": "+1. People like to complain and probably there is a reason to complain in this comp but overall it was quite interesting.
", "votes": 1, "replies": [] }, { "id": 699421, "author_name": "Dean", "author_url": "", "post_date": "2019-12-20T12:43:37.313000", "content": "I disagree as well.\nAlthough a lot of leakage data in this game, there are still more than three thousand people participate in this competition.
", "votes": 0, "replies": [] }, { "id": 699558, "author_name": "Fernando Wittmann", "author_url": "", "post_date": "2019-12-20T15:37:16.230000", "content": "\n\n\nFor example, people need to spend extra effort to collect the leakage.
\n
@naivelamb since you were not participating, probably you are not aware but some great Kagglers (@gunesevitan, @mpware, @poedator, @pdnartreb, @serengil and @pp2file) thankfully did the hard work of collecting the leakage. Also, the Kaggler @yamsam created a great kernel joining all the leakages. Thanks to them, the impact was minimal.
", "votes": 3, "replies": [] }, { "id": 699690, "author_name": "Xuan Cao", "author_url": "", "post_date": "2019-12-20T19:10:36.427000", "content": "I agree for all the competitions with leakages, there are great kagglers helping to exploit the leakages and publish it. Shout out to them. However, the question still hangs there: are all the leakages floating on the Internet being published?
\n\nI remembered that in the Google Analytics competition, when the Google Demo Account can be used to mining the target variables, lots of people left the competition and Kaggle had to reformulate it. In this competition, lots of people also left after releasing the target variables can be mined.
\n\nI think the impact of leakage can be estimate by asking one question:\n- Will you join a competition knowing there are leakages?
\n\nIf majority people answer 'No', its impact cannot be ignored.
\n\nBTW, congrats on you 18th position on public leaderboard, wish you get a good one on the private leaderboard as well.
", "votes": 0, "replies": [] } ] }, { "id": 717373, "author_name": "Clayton Miller", "author_url": "", "post_date": "2020-01-13T04:05:01.313000", "content": "Hey all! On behalf of the technical committee from ASHRAE, I just wanted to thank you all for your feedback and for sticking in there through the competition. We hope you all learned something despite the issues raised. A lot of cool analysis was created and shared through the process and we're happy for that. We learned a lot about how to host and the next one will be better, don't worry :)
", "votes": 1, "replies": [] }, { "id": 699164, "author_name": "HarshitMehta", "author_url": "", "post_date": "2019-12-20T05:41:50.737000", "content": "I believe we have seen competitions where a high scoring kernel was posted at the end or people were accused of private sharing or fake accounts (which can be true too). People will do what they wanna do, kaggle can only put regulations and do compliance to get rid of above.
\n\nThis competition differs because of leak data, but kaggle is trying to get rid of that so that it wonβt affect the final standings.
\n\nOn lighter note, I still believe that Jigsaw Toxic classification competition holds the first spot π
", "votes": 1, "replies": [] }, { "id": 699561, "author_name": "eagle4", "author_url": "", "post_date": "2019-12-20T15:38:43.627000", "content": "The worst competition ever may be this one : https://www.kaggle.com/c/draper-satellite-image-chronology/overview.
\n\nAll the gold medals, a priori, got it by manually sorting the pictures without any ML or any calculation at all.
", "votes": 2, "replies": [ { "id": 699756, "author_name": "Xuan Cao", "author_url": "", "post_date": "2019-12-20T21:23:13.053000", "content": "Luckily I wasn't playing kaggle at that time. π
", "votes": 0, "replies": [] } ] }, { "id": 699734, "author_name": "CPMP", "author_url": "", "post_date": "2019-12-20T20:35:22.703000", "content": "These issues were predictable, see for instance: https://www.kaggle.com/c/ashrae-energy-prediction/discussion/116840
", "votes": 0, "replies": [ { "id": 699755, "author_name": "Xuan Cao", "author_url": "", "post_date": "2019-12-20T21:22:28.337000", "content": "Yeah, I saw your post and decided not to join this competition.
", "votes": 1, "replies": [] } ] }, { "id": 699236, "author_name": "Manraj Singh", "author_url": "", "post_date": "2019-12-20T08:15:15.180000", "content": "'Last minute high scoring kernel.'
\n\nCan I ask where is it?
", "votes": 0, "replies": [ { "id": 699258, "author_name": "Xuan Cao", "author_url": "", "post_date": "2019-12-20T08:42:37.820000", "content": "It was mentioned in some of the threads I went through. Unfortunately, it seems to be deleted.
", "votes": 0, "replies": [] }, { "id": 699293, "author_name": "Manraj Singh", "author_url": "", "post_date": "2019-12-20T09:32:03.890000", "content": "@naivelamb I would request you to go through this discussion and kindly edit yours. :)\nhttps://www.kaggle.com/c/ashrae-energy-prediction/discussion/122446\nThank you.
", "votes": 0, "replies": [] }, { "id": 699601, "author_name": "Xuan Cao", "author_url": "", "post_date": "2019-12-20T16:18:06.590000", "content": "OK, edited.
", "votes": 1, "replies": [] } ] }, { "id": 699259, "author_name": "", "author_url": "", "post_date": "2019-12-20T08:44:04.430000", "content": "", "votes": 0, "replies": [] }, { "id": 699047, "author_name": "", "author_url": "", "post_date": "2019-12-20T02:16:51.057000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "699029": "Disclaimer: I have not entered this competition, not even looked at data.\n\nBy quickly going through the first few threads in the Discussion, it seems that this competition consists of:\n\n- Massive leakage.\n- Suspicious private sharing.\n- Suspicious fake accounts (to probe LB?).\n\nIs this the most toxic kaggle competition ever? ", "699036": "Almost 4000 teams decrease to 1800 and the number is still getting smaller.", "699225": "It was my first competition so honestly I did not know what to expect. Seeing the number of teams almost cut by half is sad. I learned many things from the amazing kernels that were shared in this competiton. \nTo all amazing kagglers, please keep sharing kernels and bring interesting insights after the end of this competiton. Thank you all !", "699046": "Honestly it's sad. It had so much potential. The data was really cool, especially because one had to learn a lot about Outlier analysis, time series and how to deal with a lot of data. It was my first competition so I learned a lot nonetheless, but I see how it could've gone so much better...", "699102": "I respectfully disagree. Here's a thread that you probably missed: https://www.kaggle.com/c/ashrae-energy-prediction/discussion/119460\n\nDespite the problems, there will be no damage in the private leaderboard and measures are being taken regarding suspicious or forbidden activities. ", "717373": "Hey all! On behalf of the technical committee from ASHRAE, I just wanted to thank you all for your feedback and for sticking in there through the competition. We hope you all learned something despite the issues raised. A lot of cool analysis was created and shared through the process and we're happy for that. We learned a lot about how to host and the next one will be better, don't worry :)", "699164": "I believe we have seen competitions where a high scoring kernel was posted at the end or people were accused of private sharing or fake accounts (which can be true too). People will do what they wanna do, kaggle can only put regulations and do compliance to get rid of above. \n\nThis competition differs because of leak data, but kaggle is trying to get rid of that so that it wonβt affect the final standings.\n\nOn lighter note, I still believe that Jigsaw Toxic classification competition holds the first spot π", "699561": "The worst competition ever may be this one : https://www.kaggle.com/c/draper-satellite-image-chronology/overview.\n\nAll the gold medals, a priori, got it by manually sorting the pictures without any ML or any calculation at all.", "699734": "These issues were predictable, see for instance: https://www.kaggle.com/c/ashrae-energy-prediction/discussion/116840", "699236": "'Last minute high scoring kernel.'\n\nCan I ask where is it?", "699259": "", "699047": "" } }