This competition is not something you see everyday on kaggle. The data is clean. The features are simple yet mysterious. The spread on leaderboard is just amazing. it is clearly the top guys are doing something really special. There are just so many things to try. So much potential! Not to mention we are predicting the stars. How cool is that!
", "messages": [ { "id": "436850", "postDate": "12/11/2018 02:54:32", "content": "This competition is not something you see everyday on kaggle. The data is clean. The features are simple yet mysterious. The spread on leaderboard is just amazing. it is clearly the top guys are doing something really special. There are just so many things to try. So much potential! Not to mention we are predicting the stars. How cool is that!
", "rawMarkdown": "This competition is not something you see everyday on kaggle. The data is clean. The features are simple yet mysterious. The spread on leaderboard is just amazing. it is clearly the top guys are doing something really special. There are just so many things to try. So much potential! Not to mention we are predicting the stars. How cool is that!", "votes": null }, { "id": "436865", "postDate": "12/11/2018 03:17:52", "content": "Although not join, I'm very curious about this amazing leaderboard and very willing to see the solutions at the end!
", "rawMarkdown": "Although not join, I'm very curious about this amazing leaderboard and very willing to see the solutions at the end!", "votes": null }, { "id": "436908", "postDate": "12/11/2018 05:03:45", "content": "Good luck everyone! I've had a lot of fun and learned a lot in this competition. I'll be very interested to see what everyone comes up with in the end.
", "rawMarkdown": "Good luck everyone! I've had a lot of fun and learned a lot in this competition. I'll be very interested to see what everyone comes up with in the end.", "votes": null }, { "id": "436928", "postDate": "12/11/2018 05:27:55", "content": "I'm sure I'l learn a lot when other teams disclose what they have done after the competition. What makes this competition special is it is open classification, with a catch all class in test. I've never worked on open classification before, and this is fascinating.
", "rawMarkdown": "I'm sure I'l learn a lot when other teams disclose what they have done after the competition. What makes this competition special is it is open classification, with a catch all class in test. I've never worked on open classification before, and this is fascinating.", "votes": null }, { "id": "436978", "postDate": "12/11/2018 07:07:42", "content": "Looking forward to everyone's solutions. Curious to know what magic leaders are doing. Do share at the end of the competition
", "rawMarkdown": "Looking forward to everyone's solutions. Curious to know what magic leaders are doing. Do share at the end of the competition", "votes": null }, { "id": "437022", "postDate": "12/11/2018 09:07:47", "content": "Yeah, great competition indeed. After spending some time on it, I feel a bit attached and wonder how can masters move from one competition to another without looking back. Is there ever a desire to continue after a competition is over?
", "rawMarkdown": "Yeah, great competition indeed. After spending some time on it, I feel a bit attached and wonder how can masters move from one competition to another without looking back. Is there ever a desire to continue after a competition is over?", "votes": null }, { "id": "437029", "postDate": "12/11/2018 09:17:26", "content": "\n\n\nIs there ever a desire to continue after a competition is over?
\n
You'll see soon that the answer is yes for many. We miss the adrenaline...
\n\nSee https://www.kaggle.com/c/quora-question-pairs/discussion/34244#189629
", "rawMarkdown": "> Is there ever a desire to continue after a competition is over?\n\nYou'll see soon that the answer is yes for many. We miss the adrenaline...\n\nSee https://www.kaggle.com/c/quora-question-pairs/discussion/34244#189629", "votes": null }, { "id": "437038", "postDate": "12/11/2018 09:34:39", "content": "the speed at which leaders are improving, it seems that it can take another 2 months and LB continue to go up up :D
", "rawMarkdown": "the speed at which leaders are improving, it seems that it can take another 2 months and LB continue to go up up :D", "votes": null }, { "id": "437041", "postDate": "12/11/2018 09:37:17", "content": "@cpmpml I mean continuing with the same competition
", "rawMarkdown": "cpmpml I mean continuing with the same competition", "votes": null }, { "id": "437051", "postDate": "12/11/2018 09:52:51", "content": "Knowing the private LB removes some of the challenge... After competition end, I usually perform some additional experiments to validate ideas I did not have time to test, but this never lasts more than few days. I guess I'm mostly motivated by the leaderboard...
", "rawMarkdown": "Knowing the private LB removes some of the challenge... After competition end, I usually perform some additional experiments to validate ideas I did not have time to test, but this never lasts more than few days. I guess I'm mostly motivated by the leaderboard...", "votes": null }, { "id": "437102", "postDate": "12/11/2018 11:31:38", "content": "Yes, I could not agree more. This competition is great and I find this is a very very tough competition. Strong variables differ in distribution in train and test, size of the test data, unknown class in test and last but not the least the problem statement (Can we help make sense of the Universe?) :D
\n\nAlthough I am new to Kaggle, I find the LB is not dense unlike the previous competitions I participated in. The leaders are doing really special work. Looking forward to their approaches and solutions.
", "rawMarkdown": "Yes, I could not agree more. This competition is great and I find this is a very very tough competition. Strong variables differ in distribution in train and test, size of the test data, unknown class in test and last but not the least the problem statement (Can we help make sense of the Universe?) :D\n\nAlthough I am new to Kaggle, I find the LB is not dense unlike the previous competitions I participated in. The leaders are doing really special work. Looking forward to their approaches and solutions.", "votes": null }, { "id": "437111", "postDate": "12/11/2018 11:46:24", "content": "Yes, very interesting data set. Despite our good score, I'm still very puzzled by many things. I'm really looking forward to seeing other solutions when the competition ends.
", "rawMarkdown": "Yes, very interesting data set. Despite our good score, I'm still very puzzled by many things. I'm really looking forward to seeing other solutions when the competition ends.", "votes": null }, { "id": "437225", "postDate": "12/11/2018 15:17:18", "content": "It's been fascinating to watch the battle between the top 3 on the leaderboard! Who will get below 0.700?!
", "rawMarkdown": "It's been fascinating to watch the battle between the top 3 on the leaderboard! Who will get below 0.700?!", "votes": null }, { "id": "437249", "postDate": "12/11/2018 16:01:17", "content": "top 4 would be fairer ;) But now it seems top 2 is even more accurate.
", "rawMarkdown": "top 4 would be fairer ;) But now it seems top 2 is even more accurate.", "votes": null }, { "id": "437253", "postDate": "12/11/2018 16:07:48", "content": "top 5 would be even more
", "rawMarkdown": "top 5 would be even more", "votes": null }, { "id": "437256", "postDate": "12/11/2018 16:09:33", "content": "@CPMP remember my LB 0.7 guess 2 months ago? ;)
", "rawMarkdown": "CPMP remember my LB 0.7 guess 2 months ago? ;)", "votes": null }, { "id": "437258", "postDate": "12/11/2018 16:11:59", "content": "I remember, and I said I believed it would be true when our best single model got 0.80 LB score ;)
", "rawMarkdown": "I remember, and I said I believed it would be true when our best single model got 0.80 LB score ;)", "votes": null }, { "id": "437272", "postDate": "12/11/2018 16:47:02", "content": "@cpmpml @yuval6967 But is this prediction correct for the right reason? You reasoned that lb score would get closer to local score as predictions for the unknown class got better. My understanding is that this is not the case and that people are trying more to improve on the classification of the known classes
", "rawMarkdown": "cpmpml @yuval6967 But is this prediction correct for the right reason? You reasoned that lb score would get closer to local score as predictions for the unknown class got better. My understanding is that this is not the case and that people are trying more to improve on the classification of the known classes", "votes": null }, { "id": "437283", "postDate": "12/11/2018 17:11:10", "content": "@CPMP
\n\nThanks a lot for all the clues and answers you gave during this competition. Really more than helpfull !
", "rawMarkdown": "CPMP\n\nThanks a lot for all the clues and answers you gave during this competition. Really more than helpfull !", "votes": null }, { "id": "437313", "postDate": "12/11/2018 17:47:04", "content": "@mezoganet, my pleasure.
\n\n@iprapas I agree with you I don't think class_99 is that critical. But that may be why top2 teams are way ahead now ;)
", "rawMarkdown": "mezoganet, my pleasure.\n\n@iprapas I agree with you I don't think class_99 is that critical. But that may be why top2 teams are way ahead now ;)", "votes": null }, { "id": "437314", "postDate": "12/11/2018 17:48:13", "content": "I am actually not enjoying this competition, I am torturing myself. I especially hate last weeks:)
", "rawMarkdown": "I am actually not enjoying this competition, I am torturing myself. I especially hate last weeks:)", "votes": null }, { "id": "437317", "postDate": "12/11/2018 17:53:55", "content": "I am especially curious about your solution Ahmet. As a single competitor and without astronomical domain -as far as I know-, your score is amazing.
", "rawMarkdown": "I am especially curious about your solution Ahmet. As a single competitor and without astronomical domain -as far as I know-, your score is amazing.", "votes": null }, { "id": "437319", "postDate": "12/11/2018 17:54:12", "content": "You have been in the fifth position for the last tow months, I guess. :) That is unbelievably cool. While the top 4 have been moving around in the first four positions. \nIt would be really interesting to see the final standing. Even though, we are nowhere close, we are enjoying this.
", "rawMarkdown": "You have been in the fifth position for the last tow months, I guess. :) That is unbelievably cool. While the top 4 have been moving around in the first four positions. \nIt would be really interesting to see the final standing. Even though, we are nowhere close, we are enjoying this.", "votes": null }, { "id": "437323", "postDate": "12/11/2018 18:01:57", "content": "I have used the name Sisyphus because normally on Kaggle, everyday I go up and down on LB. But in this competition, I might be the most stable Sisyphus ever:)
", "rawMarkdown": "I have used the name Sisyphus because normally on Kaggle, everyday I go up and down on LB. But in this competition, I might be the most stable Sisyphus ever:)", "votes": null }, { "id": "437365", "postDate": "12/11/2018 19:19:46", "content": "\n\n\nI am torturing myself
\n
First time I see torture yielding good results! ;)
", "rawMarkdown": "> I am torturing myself\n\nFirst time I see torture yielding good results! ;)", "votes": null }, { "id": "437373", "postDate": "12/11/2018 19:34:52", "content": "We're delighted so many of you enjoy this challenge! Note for those of you wondering if you should continue working after Dec. 17, there are additional science prizes you can win until Jan. 15th, and these are not evaluated using the same metric as the leaderboard because we're asking different questions. I.e., you can not be on the top of the leaderboard, but your algorithm may still be doing something that is really interesting scientifically, and we do have rewards for that!
\n\nWe'll post some of the specific questions we're interested very soon!
\n\nCheers,
\n\n-Gautham for the PLAsTiCC team
", "rawMarkdown": "We're delighted so many of you enjoy this challenge! Note for those of you wondering if you should continue working after Dec. 17, there are additional science prizes you can win until Jan. 15th, and these are not evaluated using the same metric as the leaderboard because we're asking different questions. I.e., you can not be on the top of the leaderboard, but your algorithm may still be doing something that is really interesting scientifically, and we do have rewards for that!\n\nWe'll post some of the specific questions we're interested very soon!\n\nCheers,\n\n-Gautham for the PLAsTiCC team", "votes": null }, { "id": "437429", "postDate": "12/11/2018 22:43:50", "content": "@kyle
\n\nReading your post and, ooops... you are 1st !!!
\n\nI cross my fingers for you, good luck and thank you for your comments an tips all along this trip.
", "rawMarkdown": "kyle\n\nReading your post and, ooops... you are 1st !!!\n\nI cross my fingers for you, good luck and thank you for your comments an tips all along this trip.", "votes": null }, { "id": "437430", "postDate": "12/11/2018 22:46:04", "content": "With the prize I could buy a new Celestron ?
", "rawMarkdown": "With the prize I could buy a new Celestron ?", "votes": null }, { "id": "437468", "postDate": "12/12/2018 01:27:05", "content": "@cpmpml The adrenaline... that's the narcotic that makes kaggle addictive... \n@iprapas\nI am going to have a Christmas break :)... Kaggle is dangerous for me, I got too excited... Sleepless nights, ideas, experiments, hopes, joy and disappointments, ups and downs... emotions, emotions, emotions... and lot's of learning
", "rawMarkdown": "cpmpml The adrenaline... that's the narcotic that makes kaggle addictive... \n@iprapas\nI am going to have a Christmas break :)... Kaggle is dangerous for me, I got too excited... Sleepless nights, ideas, experiments, hopes, joy and disappointments, ups and downs... emotions, emotions, emotions... and lot's of learning", "votes": null }, { "id": "437497", "postDate": "12/12/2018 03:09:05", "content": "It has been a joyride, learnt so much :D and looking to learn more from the solutions once the competition end.
", "rawMarkdown": "It has been a joyride, learnt so much :D and looking to learn more from the solutions once the competition end.", "votes": null }, { "id": "437737", "postDate": "12/12/2018 12:00:02", "content": "It was correct for wrong reasons. But still correct ;)
", "rawMarkdown": "It was correct for wrong reasons. But still correct ;)", "votes": null } ], "comments": [ { "id": 436865, "author_name": "yiheng", "author_url": "", "post_date": "12/11/2018 03:17:52", "content": "Although not join, I'm very curious about this amazing leaderboard and very willing to see the solutions at the end!
", "votes": null, "replies": [] }, { "id": 436908, "author_name": "kyleboone", "author_url": "", "post_date": "12/11/2018 05:03:45", "content": "Good luck everyone! I've had a lot of fun and learned a lot in this competition. I'll be very interested to see what everyone comes up with in the end.
", "votes": null, "replies": [ { "id": 437429, "author_name": "pfichou", "author_url": "", "post_date": "12/11/2018 22:43:50", "content": "@kyle
\n\nReading your post and, ooops... you are 1st !!!
\n\nI cross my fingers for you, good luck and thank you for your comments an tips all along this trip.
", "votes": null, "replies": [] } ] }, { "id": 436928, "author_name": "cpmpml", "author_url": "", "post_date": "12/11/2018 05:27:55", "content": "I'm sure I'l learn a lot when other teams disclose what they have done after the competition. What makes this competition special is it is open classification, with a catch all class in test. I've never worked on open classification before, and this is fascinating.
", "votes": null, "replies": [] }, { "id": 436978, "author_name": "mks2192", "author_url": "", "post_date": "12/11/2018 07:07:42", "content": "Looking forward to everyone's solutions. Curious to know what magic leaders are doing. Do share at the end of the competition
", "votes": null, "replies": [] }, { "id": 437022, "author_name": "iprapas", "author_url": "", "post_date": "12/11/2018 09:07:47", "content": "Yeah, great competition indeed. After spending some time on it, I feel a bit attached and wonder how can masters move from one competition to another without looking back. Is there ever a desire to continue after a competition is over?
", "votes": null, "replies": [ { "id": 437029, "author_name": "cpmpml", "author_url": "", "post_date": "12/11/2018 09:17:26", "content": "\n\n\nIs there ever a desire to continue after a competition is over?
\n
You'll see soon that the answer is yes for many. We miss the adrenaline...
\n\nSee https://www.kaggle.com/c/quora-question-pairs/discussion/34244#189629
", "votes": null, "replies": [] }, { "id": 437038, "author_name": "mks2192", "author_url": "", "post_date": "12/11/2018 09:34:39", "content": "the speed at which leaders are improving, it seems that it can take another 2 months and LB continue to go up up :D
", "votes": null, "replies": [] }, { "id": 437041, "author_name": "iprapas", "author_url": "", "post_date": "12/11/2018 09:37:17", "content": "@cpmpml I mean continuing with the same competition
", "votes": null, "replies": [] }, { "id": 437051, "author_name": "cpmpml", "author_url": "", "post_date": "12/11/2018 09:52:51", "content": "Knowing the private LB removes some of the challenge... After competition end, I usually perform some additional experiments to validate ideas I did not have time to test, but this never lasts more than few days. I guess I'm mostly motivated by the leaderboard...
", "votes": null, "replies": [] }, { "id": 437468, "author_name": "blondinka", "author_url": "", "post_date": "12/12/2018 01:27:05", "content": "@cpmpml The adrenaline... that's the narcotic that makes kaggle addictive... \n@iprapas\nI am going to have a Christmas break :)... Kaggle is dangerous for me, I got too excited... Sleepless nights, ideas, experiments, hopes, joy and disappointments, ups and downs... emotions, emotions, emotions... and lot's of learning
", "votes": null, "replies": [] } ] }, { "id": 437102, "author_name": "indranilbhattacharya", "author_url": "", "post_date": "12/11/2018 11:31:38", "content": "Yes, I could not agree more. This competition is great and I find this is a very very tough competition. Strong variables differ in distribution in train and test, size of the test data, unknown class in test and last but not the least the problem statement (Can we help make sense of the Universe?) :D
\n\nAlthough I am new to Kaggle, I find the LB is not dense unlike the previous competitions I participated in. The leaders are doing really special work. Looking forward to their approaches and solutions.
", "votes": null, "replies": [] }, { "id": 437111, "author_name": "psilogram", "author_url": "", "post_date": "12/11/2018 11:46:24", "content": "Yes, very interesting data set. Despite our good score, I'm still very puzzled by many things. I'm really looking forward to seeing other solutions when the competition ends.
", "votes": null, "replies": [] }, { "id": 437225, "author_name": "sdoria", "author_url": "", "post_date": "12/11/2018 15:17:18", "content": "It's been fascinating to watch the battle between the top 3 on the leaderboard! Who will get below 0.700?!
", "votes": null, "replies": [ { "id": 437249, "author_name": "cpmpml", "author_url": "", "post_date": "12/11/2018 16:01:17", "content": "top 4 would be fairer ;) But now it seems top 2 is even more accurate.
", "votes": null, "replies": [] }, { "id": 437253, "author_name": "mks2192", "author_url": "", "post_date": "12/11/2018 16:07:48", "content": "top 5 would be even more
", "votes": null, "replies": [] }, { "id": 437256, "author_name": "yuval6967", "author_url": "", "post_date": "12/11/2018 16:09:33", "content": "@CPMP remember my LB 0.7 guess 2 months ago? ;)
", "votes": null, "replies": [] }, { "id": 437258, "author_name": "cpmpml", "author_url": "", "post_date": "12/11/2018 16:11:59", "content": "I remember, and I said I believed it would be true when our best single model got 0.80 LB score ;)
", "votes": null, "replies": [] }, { "id": 437272, "author_name": "iprapas", "author_url": "", "post_date": "12/11/2018 16:47:02", "content": "@cpmpml @yuval6967 But is this prediction correct for the right reason? You reasoned that lb score would get closer to local score as predictions for the unknown class got better. My understanding is that this is not the case and that people are trying more to improve on the classification of the known classes
", "votes": null, "replies": [] }, { "id": 437283, "author_name": "pfichou", "author_url": "", "post_date": "12/11/2018 17:11:10", "content": "@CPMP
\n\nThanks a lot for all the clues and answers you gave during this competition. Really more than helpfull !
", "votes": null, "replies": [] }, { "id": 437313, "author_name": "cpmpml", "author_url": "", "post_date": "12/11/2018 17:47:04", "content": "@mezoganet, my pleasure.
\n\n@iprapas I agree with you I don't think class_99 is that critical. But that may be why top2 teams are way ahead now ;)
", "votes": null, "replies": [] }, { "id": 437737, "author_name": "yuval6967", "author_url": "", "post_date": "12/12/2018 12:00:02", "content": "It was correct for wrong reasons. But still correct ;)
", "votes": null, "replies": [] } ] }, { "id": 437314, "author_name": "aerdem4", "author_url": "", "post_date": "12/11/2018 17:48:13", "content": "I am actually not enjoying this competition, I am torturing myself. I especially hate last weeks:)
", "votes": null, "replies": [ { "id": 437317, "author_name": "fatihozturk", "author_url": "", "post_date": "12/11/2018 17:53:55", "content": "I am especially curious about your solution Ahmet. As a single competitor and without astronomical domain -as far as I know-, your score is amazing.
", "votes": null, "replies": [] }, { "id": 437319, "author_name": "indranilbhattacharya", "author_url": "", "post_date": "12/11/2018 17:54:12", "content": "You have been in the fifth position for the last tow months, I guess. :) That is unbelievably cool. While the top 4 have been moving around in the first four positions. \nIt would be really interesting to see the final standing. Even though, we are nowhere close, we are enjoying this.
", "votes": null, "replies": [] }, { "id": 437323, "author_name": "aerdem4", "author_url": "", "post_date": "12/11/2018 18:01:57", "content": "I have used the name Sisyphus because normally on Kaggle, everyday I go up and down on LB. But in this competition, I might be the most stable Sisyphus ever:)
", "votes": null, "replies": [] }, { "id": 437365, "author_name": "cpmpml", "author_url": "", "post_date": "12/11/2018 19:19:46", "content": "\n\n\nI am torturing myself
\n
First time I see torture yielding good results! ;)
", "votes": null, "replies": [] } ] }, { "id": 437373, "author_name": "gsnarayan", "author_url": "", "post_date": "12/11/2018 19:34:52", "content": "We're delighted so many of you enjoy this challenge! Note for those of you wondering if you should continue working after Dec. 17, there are additional science prizes you can win until Jan. 15th, and these are not evaluated using the same metric as the leaderboard because we're asking different questions. I.e., you can not be on the top of the leaderboard, but your algorithm may still be doing something that is really interesting scientifically, and we do have rewards for that!
\n\nWe'll post some of the specific questions we're interested very soon!
\n\nCheers,
\n\n-Gautham for the PLAsTiCC team
", "votes": null, "replies": [ { "id": 437430, "author_name": "pfichou", "author_url": "", "post_date": "12/11/2018 22:46:04", "content": "With the prize I could buy a new Celestron ?
", "votes": null, "replies": [] } ] }, { "id": 437497, "author_name": "rahulsrk", "author_url": "", "post_date": "12/12/2018 03:09:05", "content": "It has been a joyride, learnt so much :D and looking to learn more from the solutions once the competition end.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "436850": "This competition is not something you see everyday on kaggle. The data is clean. The features are simple yet mysterious. The spread on leaderboard is just amazing. it is clearly the top guys are doing something really special. There are just so many things to try. So much potential! Not to mention we are predicting the stars. How cool is that!", "436865": "Although not join, I'm very curious about this amazing leaderboard and very willing to see the solutions at the end!", "436908": "Good luck everyone! I've had a lot of fun and learned a lot in this competition. I'll be very interested to see what everyone comes up with in the end.", "436928": "I'm sure I'l learn a lot when other teams disclose what they have done after the competition. What makes this competition special is it is open classification, with a catch all class in test. I've never worked on open classification before, and this is fascinating.", "436978": "Looking forward to everyone's solutions. Curious to know what magic leaders are doing. Do share at the end of the competition", "437022": "Yeah, great competition indeed. After spending some time on it, I feel a bit attached and wonder how can masters move from one competition to another without looking back. Is there ever a desire to continue after a competition is over?", "437029": "> Is there ever a desire to continue after a competition is over?\n\nYou'll see soon that the answer is yes for many. We miss the adrenaline...\n\nSee https://www.kaggle.com/c/quora-question-pairs/discussion/34244#189629", "437038": "the speed at which leaders are improving, it seems that it can take another 2 months and LB continue to go up up :D", "437041": "cpmpml I mean continuing with the same competition", "437051": "Knowing the private LB removes some of the challenge... After competition end, I usually perform some additional experiments to validate ideas I did not have time to test, but this never lasts more than few days. I guess I'm mostly motivated by the leaderboard...", "437102": "Yes, I could not agree more. This competition is great and I find this is a very very tough competition. Strong variables differ in distribution in train and test, size of the test data, unknown class in test and last but not the least the problem statement (Can we help make sense of the Universe?) :D\n\nAlthough I am new to Kaggle, I find the LB is not dense unlike the previous competitions I participated in. The leaders are doing really special work. Looking forward to their approaches and solutions.", "437111": "Yes, very interesting data set. Despite our good score, I'm still very puzzled by many things. I'm really looking forward to seeing other solutions when the competition ends.", "437225": "It's been fascinating to watch the battle between the top 3 on the leaderboard! Who will get below 0.700?!", "437249": "top 4 would be fairer ;) But now it seems top 2 is even more accurate.", "437253": "top 5 would be even more", "437256": "CPMP remember my LB 0.7 guess 2 months ago? ;)", "437258": "I remember, and I said I believed it would be true when our best single model got 0.80 LB score ;)", "437272": "cpmpml @yuval6967 But is this prediction correct for the right reason? You reasoned that lb score would get closer to local score as predictions for the unknown class got better. My understanding is that this is not the case and that people are trying more to improve on the classification of the known classes", "437283": "CPMP\n\nThanks a lot for all the clues and answers you gave during this competition. Really more than helpfull !", "437313": "mezoganet, my pleasure.\n\n@iprapas I agree with you I don't think class_99 is that critical. But that may be why top2 teams are way ahead now ;)", "437314": "I am actually not enjoying this competition, I am torturing myself. I especially hate last weeks:)", "437317": "I am especially curious about your solution Ahmet. As a single competitor and without astronomical domain -as far as I know-, your score is amazing.", "437319": "You have been in the fifth position for the last tow months, I guess. :) That is unbelievably cool. While the top 4 have been moving around in the first four positions. \nIt would be really interesting to see the final standing. Even though, we are nowhere close, we are enjoying this.", "437323": "I have used the name Sisyphus because normally on Kaggle, everyday I go up and down on LB. But in this competition, I might be the most stable Sisyphus ever:)", "437365": "> I am torturing myself\n\nFirst time I see torture yielding good results! ;)", "437373": "We're delighted so many of you enjoy this challenge! Note for those of you wondering if you should continue working after Dec. 17, there are additional science prizes you can win until Jan. 15th, and these are not evaluated using the same metric as the leaderboard because we're asking different questions. I.e., you can not be on the top of the leaderboard, but your algorithm may still be doing something that is really interesting scientifically, and we do have rewards for that!\n\nWe'll post some of the specific questions we're interested very soon!\n\nCheers,\n\n-Gautham for the PLAsTiCC team", "437429": "kyle\n\nReading your post and, ooops... you are 1st !!!\n\nI cross my fingers for you, good luck and thank you for your comments an tips all along this trip.", "437430": "With the prize I could buy a new Celestron ?", "437468": "cpmpml The adrenaline... that's the narcotic that makes kaggle addictive... \n@iprapas\nI am going to have a Christmas break :)... Kaggle is dangerous for me, I got too excited... Sleepless nights, ideas, experiments, hopes, joy and disappointments, ups and downs... emotions, emotions, emotions... and lot's of learning", "437497": "It has been a joyride, learnt so much :D and looking to learn more from the solutions once the competition end.", "437737": "It was correct for wrong reasons. But still correct ;)" }, "source": "meta" }