That's strange!
\n\n212903,DagnyT,"2015-09-11 07:19:46",0.99805\n207372,DataMinders,"2015-09-11 09:57:55",0.99805\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!
", "rawMarkdown": "Sounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!", "votes": 9 }, { "id": 92175, "postDate": "2015-09-11T17:21:00.713Z", "content": "[quote=Bluefool;92124]\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!\n[/quote]
\n\nA W E S O M E!!!, Thanks :D
\n\nAlthough the two perfect scores makes me think this could end up with Top 10 perfect scores that won't include me XD. None of the those teams have shared any external data usage, so... there must be some magic features indeed.
\n\nBack on topic, the Kaggle rules regarding private sharing are very clear, but its none of us job to enforce them, so just relax, if required, it will happen.
", "rawMarkdown": "[quote=Bluefool;92124]\r\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!\r\n[/quote]\r\n\r\nA W E S O M E!!!, Thanks :D\r\n\r\nAlthough the two perfect scores makes me think this could end up with Top 10 perfect scores that won't include me XD. None of the those teams have [shared][1] any external data usage, so... there must be some magic features indeed.\r\n\r\nBack on topic, the Kaggle rules regarding private sharing are very clear, but its none of us job to enforce them, so just relax, if required, it will happen.\r\n\r\n\r\n [1]: https://www.kaggle.com/c/dato-native/forums/t/16272/external-data-repository", "votes": 5 }, { "id": 92127, "postDate": "2015-09-11T13:08:51.683Z", "content": "[quote=Bluefool;92124]
\n\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!
\n\n[/quote]
\n\nI don't need to mention that you are one of the few people who always say fair things :)
", "rawMarkdown": "[quote=Bluefool;92124]\r\n\r\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!\r\n\r\n[/quote]\r\n\r\nI don't need to mention that you are one of the few people who always say fair things :)", "votes": 6 }, { "id": 92118, "postDate": "2015-09-11T12:44:38.670Z", "content": "[quote=rcarson;92115]
\n\nTwo same score and a quick merge. Did you plan to merge after you two get the same score or before it?\nI am in no position to judge it but I doubt if you two achieve that score independently.
\n\n[/quote]
\n\nLet me clarify few things before people start running there minds in all direction
\n\nDagnyT is the third member of DataMinders team (you can check past competition) and decided to start working on competition independently ( I don't know how long DagnyT started). Few hours back, DagnyT after submitting best model of current score shared the intention to form team and asked Oleksii to merge team. It just happen to be the case I was the Leader (Kaggle can confirm that) and Oleksii couldn't send the invite and I was busy in the office and got to know it later. They decided to merge repositories by the time I come back and form the team as soon as I see Oleksii message on hangout. It just happen to be the case Oleksii was testing the model with different seed and ended up submitting and getting the same score. Also Kaggle can confirm that we only made 3 submission in total including before and after the formation of team and best model was submitted today only. Also, I made Oleksii leader after seeing the message and then he formed the team.
\n\nLet me make it clear what you were trying to say, DID YOU GUYS CHEAT? \n- answer is NO WE DIDN'T
\n\nSorry If I am being rude but I just don't like people starting to run their mind without knowing what exactly happened.
", "rawMarkdown": "[quote=rcarson;92115]\r\n\r\nTwo same score and a quick merge. Did you plan to merge after you two get the same score or before it?\r\nI am in no position to judge it but I doubt if you two achieve that score independently.\r\n\r\n[/quote]\r\n\r\nLet me clarify few things before people start running there minds in all direction\r\n\r\nDagnyT is the third member of DataMinders team (you can check past competition) and decided to start working on competition independently ( I don't know how long DagnyT started). Few hours back, DagnyT after submitting best model of current score shared the intention to form team and asked Oleksii to merge team. It just happen to be the case I was the Leader (Kaggle can confirm that) and Oleksii couldn't send the invite and I was busy in the office and got to know it later. They decided to merge repositories by the time I come back and form the team as soon as I see Oleksii message on hangout. It just happen to be the case Oleksii was testing the model with different seed and ended up submitting and getting the same score. Also Kaggle can confirm that we only made 3 submission in total including before and after the formation of team and best model was submitted today only. Also, I made Oleksii leader after seeing the message and then he formed the team.\r\n\r\nLet me make it clear what you were trying to say, DID YOU GUYS CHEAT? \r\n- answer is NO WE DIDN'T\r\n\r\nSorry If I am being rude but I just don't like people starting to run their mind without knowing what exactly happened.", "votes": 5 }, { "id": 92160, "postDate": "2015-09-11T16:01:34.543Z", "content": "I would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)
", "rawMarkdown": "I would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)", "votes": 4 }, { "id": 92164, "postDate": "2015-09-11T16:14:43.860Z", "content": "[quote=Little Boat;92160]
\n\nI would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)
\n\n[/quote]
\n\nHey, I am not the victim and this is just a thread that everybody earns some upvotes : D
", "rawMarkdown": "[quote=Little Boat;92160]\r\n\r\nI would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)\r\n\r\n[/quote]\r\n\r\nHey, I am not the victim and this is just a thread that everybody earns some upvotes : D", "votes": 5 }, { "id": 92093, "postDate": "2015-09-11T10:19:06.507Z", "content": "That's strange!
\n\n212903,DagnyT,"2015-09-11 07:19:46",0.99805\n207372,DataMinders,"2015-09-11 09:57:55",0.99805\nWhen I first saw the almost perfect score I thought there had to be something deterministic. I don't think there will be a perfect score though. If you look at very small files, there are some that are identical but one is a 1 and the other is a 0. At this point it would make sense to just patiently sift through some larger files and manually find small differences that are not captured by the standard features most people are using. For instance, I was looking for "insignificant" differences something as little as presence/absence of a space before an = sign in a tag, that might differentiate the underlying code generation method of native ads pages (or whatever StumbleUpon is calling native ads pages). Anyway, I really hope to be wrong, but this doesn't feel like a ML exercise at this point.
", "rawMarkdown": "When I first saw the almost perfect score I thought there had to be something deterministic. I don't think there will be a perfect score though. If you look at very small files, there are some that are identical but one is a 1 and the other is a 0. At this point it would make sense to just patiently sift through some larger files and manually find small differences that are not captured by the standard features most people are using. For instance, I was looking for \"insignificant\" differences something as little as presence/absence of a space before an = sign in a tag, that might differentiate the underlying code generation method of native ads pages (or whatever StumbleUpon is calling native ads pages). Anyway, I really hope to be wrong, but this doesn't feel like a ML exercise at this point. ", "votes": 4 }, { "id": 92132, "postDate": "2015-09-11T13:55:27.577Z", "content": "[quote=DataGeek;92126]
\n\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.
\n\n[/quote]
\n\nI am speaking my mind. I don't say you cheat. I don't mean it at all. \nBut the fact was that "Oleksii was testing the model (DagnyT's model?) with different seed and ended up submitting and getting the same score", when Oleksii and DagnyT were in two different teams. Yes, it was a mistake. But was it private sharing? When we talk about private sharing, does it have to do with just facts, or mistakes and intentions?
\n\nI'm no one and I don't think your mistakes need to be punished since it hurts no one. But if anyone "trying to make seen out of it", it's not me. Mistakes did get people removed, https://www.kaggle.com/c/seizure-prediction/forums/t/10972/the-one-with-only-21-entries-removed
\n\nI hope we all won't make such mistakes.
", "rawMarkdown": "[quote=DataGeek;92126]\r\n\r\n\r\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.\r\n\r\n\r\n[/quote]\r\n\r\nI am speaking my mind. I don't say you cheat. I don't mean it at all. \r\nBut the fact was that \"Oleksii was testing the model (DagnyT's model?) with different seed and ended up submitting and getting the same score\", when Oleksii and DagnyT were in two different teams. Yes, it was a mistake. But was it private sharing? When we talk about private sharing, does it have to do with just facts, or mistakes and intentions?\r\n\r\nI'm no one and I don't think your mistakes need to be punished since it hurts no one. But if anyone \"trying to make seen out of it\", it's not me. Mistakes did get people removed, https://www.kaggle.com/c/seizure-prediction/forums/t/10972/the-one-with-only-21-entries-removed\r\n\r\nI hope we all won't make such mistakes.\r\n", "votes": 2 }, { "id": 92126, "postDate": "2015-09-11T13:07:44.533Z", "content": "[quote=rcarson;92123]
\n\n[quote=DataGeek;92121]
\n\n[quote=rcarson;92115]
\n\nEdit: Basically I am against collaboration before merging.
\n\n[/quote]
\n\nI have explained that pretty well and let me know if you still have any doubts.
\n\n[/quote]
\n\nThank you for explaining. What I feel doesn't matter, it is up to the admin to decide if this violates the 'no private sharing rule' or not.
\n\n[/quote]
\n\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.
", "rawMarkdown": "[quote=rcarson;92123]\r\n\r\n[quote=DataGeek;92121]\r\n\r\n[quote=rcarson;92115]\r\n\r\nEdit: Basically I am against collaboration before merging.\r\n\r\n[/quote]\r\n\r\nI have explained that pretty well and let me know if you still have any doubts.\r\n\r\n[/quote]\r\n\r\nThank you for explaining. What I feel doesn't matter, it is up to the admin to decide if this violates the 'no private sharing rule' or not.\r\n\r\n[/quote]\r\n\r\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.\r\n", "votes": 2 }, { "id": 92143, "postDate": "2015-09-11T14:25:43.677Z", "content": "@rcarson I'm not scared of this mistake, then why I have to hide it? A good point to learn something. It wasn't cheating. It was a mistake. And about private sharing, if it's needed, for sure, I'll explain exactly what happened to Admin and Admin then will decide.
", "rawMarkdown": "@rcarson I'm not scared of this mistake, then why I have to hide it? A good point to learn something. It wasn't cheating. It was a mistake. And about private sharing, if it's needed, for sure, I'll explain exactly what happened to Admin and Admin then will decide. ", "votes": 1 }, { "id": 92114, "postDate": "2015-09-11T12:29:16.750Z", "content": "[quote=rcarson;92113]
\n\nThis feels wrong
\n\n[/quote]
\n\nWhat exactly feels wrong ?
", "rawMarkdown": "[quote=rcarson;92113]\r\n\r\nThis feels wrong\r\n\r\n[/quote]\r\n\r\nWhat exactly feels wrong ?", "votes": 1 }, { "id": 92115, "postDate": "2015-09-11T12:32:31.943Z", "content": "Two same score and a quick merge. Did you plan to merge after you two get the same score or before it?\nI am in no position to judge it but I doubt if you two achieve that score independently.
\n\nEdit: Basically I am against collaboration before merging.
", "rawMarkdown": "Two same score and a quick merge. Did you plan to merge after you two get the same score or before it?\r\nI am in no position to judge it but I doubt if you two achieve that score independently.\r\n\r\nEdit: Basically I am against collaboration before merging.", "votes": 2 }, { "id": 92173, "postDate": "2015-09-11T17:17:57.180Z", "content": "[quote=rcarson;92142]
\n\n@ DataGeek, I want to apologize if I sound offensive. I hope we could do as well as your team.
\n\nGood luck!
\n\n[/quote]
\n\nI think you were right at your place and was just trying to get the details. I am sorry for being rude.
", "rawMarkdown": "[quote=rcarson;92142]\r\n\r\n@ DataGeek, I want to apologize if I sound offensive. I hope we could do as well as your team. \r\n\r\nGood luck!\r\n\r\n[/quote]\r\n\r\nI think you were right at your place and was just trying to get the details. I am sorry for being rude.\r\n", "votes": 1 }, { "id": 92121, "postDate": "2015-09-11T12:52:52.780Z", "content": "[quote=rcarson;92115]
\n\nEdit: Basically I am against collaboration before merging.
\n\n[/quote]
\n\nI have explained that pretty well and let me know if you still have any doubts.
", "rawMarkdown": "[quote=rcarson;92115]\r\n\r\nEdit: Basically I am against collaboration before merging.\r\n\r\n[/quote]\r\n\r\nI have explained that pretty well and let me know if you still have any doubts.", "votes": 1 }, { "id": 92202, "postDate": "2015-09-11T20:51:29.203Z", "content": "Make it 3 new perfect scores....
", "rawMarkdown": "Make it 3 new perfect scores....", "votes": 2 }, { "id": 92177, "postDate": "2015-09-11T17:27:52.673Z", "content": "\n\n\nAlthough the two perfect scores makes me think this could end up with Top 10 perfect scores
\n
I assumed it with the one score, but now as more come in very much think so, that there is likely some deterministic or leakage thing here, and then the end will be arrow splitting potentially off of the "other" model the leakage is mixed with (particularly if they redo the data and then we go back to the traditional models).
\n\nI saw something that I suspected may be what they are doing, but my code hasn't finished running yet, so not sure if it is real or apophenia at this point - will see this weekend I guess.
", "rawMarkdown": "> Although the two perfect scores makes me think this could end up with Top 10 perfect scores\r\n\r\nI assumed it with the one score, but now as more come in very much think so, that there is likely some deterministic or leakage thing here, and then the end will be arrow splitting potentially off of the \"other\" model the leakage is mixed with (particularly if they redo the data and then we go back to the traditional models).\r\n\r\nI saw something that I suspected may be what they are doing, but my code hasn't finished running yet, so not sure if it is real or apophenia at this point - will see this weekend I guess.", "votes": 2 }, { "id": 92142, "postDate": "2015-09-11T14:22:52.823Z", "content": "@ DataGeek, I want to apologize if I sound offensive. I hope we could do as well as your team.
\n\nGood luck!
", "rawMarkdown": "@ DataGeek, I want to apologize if I sound offensive. I hope we could do as well as your team. \r\n\r\nGood luck!", "votes": 2 }, { "id": 92141, "postDate": "2015-09-11T14:21:34.590Z", "content": "[quote=DagnyT;92138]
\n\nKeyword - sometimes :D , it was a really very bad mistake, worth discussion? Or you just want us to be punished for it?
\n\n[/quote]
\n\nHey, I don't want you to be punished. I am just discussing what is 'private sharing' with DataGeek, and maybe everyone else.
\n\nFor me, it's Ok to make mistakes as long as people don't find out. : )
", "rawMarkdown": "[quote=DagnyT;92138]\r\n\r\nKeyword - sometimes :D , it was a really very bad mistake, worth discussion? Or you just want us to be punished for it? \r\n\r\n[/quote]\r\n\r\nHey, I don't want you to be punished. I am just discussing what is 'private sharing' with DataGeek, and maybe everyone else. \r\n\r\nFor me, it's Ok to make mistakes as long as people don't find out. : )" }, { "id": 92137, "postDate": "2015-09-11T14:10:42.300Z", "content": "[quote=DagnyT;92136]
\n\n@rcarson I really hope you don't make any mistake in your life, ideal one. Have a great day!
\n\n[/quote]
\n\nI make mistakes everyday and I am punished sometimes. : )
", "rawMarkdown": "[quote=DagnyT;92136]\r\n\r\n@rcarson I really hope you don't make any mistake in your life, ideal one. Have a great day! \r\n\r\n[/quote]\r\n\r\nI make mistakes everyday and I am punished sometimes. : )" }, { "id": 92119, "postDate": "2015-09-11T12:45:28.550Z", "content": "[quote=rcarson;92115]
\n\nTwo same score and a quick merge. Did you plan to merge after you two get the same score or before it?\nI am in no position to judge it but I doubt if you two achieve that score independently.
\n\nEdit: Basically I am against collaboration before merging.
\n\n[/quote]
\n\nTotally agree, if they didn't collaborate before merging, I don't think they have the necessity to collaborate at this time.
", "rawMarkdown": "[quote=rcarson;92115]\r\n\r\nTwo same score and a quick merge. Did you plan to merge after you two get the same score or before it?\r\nI am in no position to judge it but I doubt if you two achieve that score independently.\r\n\r\nEdit: Basically I am against collaboration before merging.\r\n\r\n[/quote]\r\n\r\nTotally agree, if they didn't collaborate before merging, I don't think they have the necessity to collaborate at this time." }, { "id": 92139, "postDate": "2015-09-11T14:17:08.437Z", "content": "[quote=rcarson;92132]
\n\n[quote=DataGeek;92126]
\n\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.
\n\n[/quote]
\n\nI am speaking my mind. I don't say you cheat. I don't mean it at all. \nBut the fact was that "Oleksii was testing the model (DagnyT's model?) with different seed and ended up submitting and getting the same score", when Oleksii and DagnyT were in two different teams. Yes, it was a mistake. But was it private sharing? When we talk about private sharing, does it have to do with just facts, or mistakes and intentions?
\n\nI'm no one and I don't think your mistakes need to be punished since it hurts no one. But if anyone "trying to make seen out of it", it's not me. Mistakes did get people removed, https://www.kaggle.com/c/seizure-prediction/forums/t/10972/the-one-with-only-21-entries-removed
\n\nI hope we all won't make such mistakes.
\n\n[/quote]
\n\nI agree with you and assure you there was no intention to cheat and no cheating has been done. We didn't try to use extra submission and you/admin can cross verify that with Kaggle that we only made 3 submission today. Also, admin can check all other previous submission of DagnyT and Our. It's our honesty that we explained what happened exactly. If admin finds we have done something wrong admins can remove us.
", "rawMarkdown": "[quote=rcarson;92132]\r\n\r\n[quote=DataGeek;92126]\r\n\r\n\r\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.\r\n\r\n\r\n[/quote]\r\n\r\nI am speaking my mind. I don't say you cheat. I don't mean it at all. \r\nBut the fact was that \"Oleksii was testing the model (DagnyT's model?) with different seed and ended up submitting and getting the same score\", when Oleksii and DagnyT were in two different teams. Yes, it was a mistake. But was it private sharing? When we talk about private sharing, does it have to do with just facts, or mistakes and intentions?\r\n\r\nI'm no one and I don't think your mistakes need to be punished since it hurts no one. But if anyone \"trying to make seen out of it\", it's not me. Mistakes did get people removed, https://www.kaggle.com/c/seizure-prediction/forums/t/10972/the-one-with-only-21-entries-removed\r\n\r\nI hope we all won't make such mistakes.\r\n\r\n\r\n[/quote]\r\n\r\nI agree with you and assure you there was no intention to cheat and no cheating has been done. We didn't try to use extra submission and you/admin can cross verify that with Kaggle that we only made 3 submission today. Also, admin can check all other previous submission of DagnyT and Our. It's our honesty that we explained what happened exactly. If admin finds we have done something wrong admins can remove us.\r\n", "votes": 1 }, { "id": 92105, "postDate": "2015-09-11T11:35:37.887Z", "content": "They probably merged afterwards.
\n\nI still have a window open with it.
\n\nHere is the screen capture.
", "rawMarkdown": "They probably merged afterwards.\r\n\r\nI still have a window open with it.\r\n\r\nHere is the screen capture. ", "votes": 1 }, { "id": 92113, "postDate": "2015-09-11T12:26:20.803Z", "content": "This feels wrong
", "rawMarkdown": "This feels wrong" }, { "id": 92120, "postDate": "2015-09-11T12:45:36.533Z", "rawMarkdown": "", "votes": -1 }, { "id": 92174, "postDate": "2015-09-11T17:20:34.027Z", "content": "[quote=rcarson;92164]
\n\n[quote=Little Boat;92160]
\n\nI would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)
\n\n[/quote]
\n\nHey, I am not the victim and this is just a thread that everybody earns some upvotes : D
\n\n[/quote]
\n\n+1
", "rawMarkdown": "[quote=rcarson;92164]\r\n\r\n[quote=Little Boat;92160]\r\n\r\nI would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)\r\n\r\n[/quote]\r\n\r\nHey, I am not the victim and this is just a thread that everybody earns some upvotes : D\r\n\r\n[/quote]\r\n\r\n+1" }, { "id": 92123, "postDate": "2015-09-11T12:58:59.927Z", "content": "[quote=DataGeek;92121]
\n\n[quote=rcarson;92115]
\n\nEdit: Basically I am against collaboration before merging.
\n\n[/quote]
\n\nI have explained that pretty well and let me know if you still have any doubts.
\n\n[/quote]
\n\nThank you for explaining. What I feel doesn't matter, it is up to the admin to decide if this violates the 'no private sharing rule' or not.
", "rawMarkdown": "[quote=DataGeek;92121]\r\n\r\n[quote=rcarson;92115]\r\n\r\nEdit: Basically I am against collaboration before merging.\r\n\r\n[/quote]\r\n\r\nI have explained that pretty well and let me know if you still have any doubts.\r\n\r\n[/quote]\r\n\r\nThank you for explaining. What I feel doesn't matter, it is up to the admin to decide if this violates the 'no private sharing rule' or not." }, { "id": 92094, "postDate": "2015-09-11T10:41:31.467Z", "content": "I didn't see that, the merged together.
", "rawMarkdown": "I didn't see that, the merged together." }, { "id": 92138, "postDate": "2015-09-11T14:13:22.687Z", "rawMarkdown": "", "votes": -3, "isDeleted": true }, { "id": 92136, "postDate": "2015-09-11T14:09:23.090Z", "rawMarkdown": "", "votes": -5, "isDeleted": true } ], "comments": [ { "id": 92124, "author_name": "Bluefool", "author_url": "", "post_date": "2015-09-11T13:03:20.857000", "content": "Sounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!
", "votes": 9, "replies": [] }, { "id": 92175, "author_name": "NxGTR", "author_url": "", "post_date": "2015-09-11T17:21:00.713000", "content": "[quote=Bluefool;92124]\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!\n[/quote]
\n\nA W E S O M E!!!, Thanks :D
\n\nAlthough the two perfect scores makes me think this could end up with Top 10 perfect scores that won't include me XD. None of the those teams have shared any external data usage, so... there must be some magic features indeed.
\n\nBack on topic, the Kaggle rules regarding private sharing are very clear, but its none of us job to enforce them, so just relax, if required, it will happen.
", "votes": 5, "replies": [] }, { "id": 92127, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T13:08:51.683000", "content": "[quote=Bluefool;92124]
\n\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!
\n\n[/quote]
\n\nI don't need to mention that you are one of the few people who always say fair things :)
", "votes": 6, "replies": [] }, { "id": 92118, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T12:44:38.670000", "content": "[quote=rcarson;92115]
\n\nTwo same score and a quick merge. Did you plan to merge after you two get the same score or before it?\nI am in no position to judge it but I doubt if you two achieve that score independently.
\n\n[/quote]
\n\nLet me clarify few things before people start running there minds in all direction
\n\nDagnyT is the third member of DataMinders team (you can check past competition) and decided to start working on competition independently ( I don't know how long DagnyT started). Few hours back, DagnyT after submitting best model of current score shared the intention to form team and asked Oleksii to merge team. It just happen to be the case I was the Leader (Kaggle can confirm that) and Oleksii couldn't send the invite and I was busy in the office and got to know it later. They decided to merge repositories by the time I come back and form the team as soon as I see Oleksii message on hangout. It just happen to be the case Oleksii was testing the model with different seed and ended up submitting and getting the same score. Also Kaggle can confirm that we only made 3 submission in total including before and after the formation of team and best model was submitted today only. Also, I made Oleksii leader after seeing the message and then he formed the team.
\n\nLet me make it clear what you were trying to say, DID YOU GUYS CHEAT? \n- answer is NO WE DIDN'T
\n\nSorry If I am being rude but I just don't like people starting to run their mind without knowing what exactly happened.
", "votes": 5, "replies": [] }, { "id": 92160, "author_name": "Little Boat", "author_url": "", "post_date": "2015-09-11T16:01:34.543000", "content": "I would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)
", "votes": 4, "replies": [] }, { "id": 92164, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2015-09-11T16:14:43.860000", "content": "[quote=Little Boat;92160]
\n\nI would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)
\n\n[/quote]
\n\nHey, I am not the victim and this is just a thread that everybody earns some upvotes : D
", "votes": 5, "replies": [] }, { "id": 92185, "author_name": "Giulio", "author_url": "", "post_date": "2015-09-11T18:03:39.603000", "content": "When I first saw the almost perfect score I thought there had to be something deterministic. I don't think there will be a perfect score though. If you look at very small files, there are some that are identical but one is a 1 and the other is a 0. At this point it would make sense to just patiently sift through some larger files and manually find small differences that are not captured by the standard features most people are using. For instance, I was looking for "insignificant" differences something as little as presence/absence of a space before an = sign in a tag, that might differentiate the underlying code generation method of native ads pages (or whatever StumbleUpon is calling native ads pages). Anyway, I really hope to be wrong, but this doesn't feel like a ML exercise at this point.
", "votes": 4, "replies": [] }, { "id": 92132, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2015-09-11T13:55:27.577000", "content": "[quote=DataGeek;92126]
\n\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.
\n\n[/quote]
\n\nI am speaking my mind. I don't say you cheat. I don't mean it at all. \nBut the fact was that "Oleksii was testing the model (DagnyT's model?) with different seed and ended up submitting and getting the same score", when Oleksii and DagnyT were in two different teams. Yes, it was a mistake. But was it private sharing? When we talk about private sharing, does it have to do with just facts, or mistakes and intentions?
\n\nI'm no one and I don't think your mistakes need to be punished since it hurts no one. But if anyone "trying to make seen out of it", it's not me. Mistakes did get people removed, https://www.kaggle.com/c/seizure-prediction/forums/t/10972/the-one-with-only-21-entries-removed
\n\nI hope we all won't make such mistakes.
", "votes": 2, "replies": [] }, { "id": 92126, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T13:07:44.533000", "content": "[quote=rcarson;92123]
\n\n[quote=DataGeek;92121]
\n\n[quote=rcarson;92115]
\n\nEdit: Basically I am against collaboration before merging.
\n\n[/quote]
\n\nI have explained that pretty well and let me know if you still have any doubts.
\n\n[/quote]
\n\nThank you for explaining. What I feel doesn't matter, it is up to the admin to decide if this violates the 'no private sharing rule' or not.
\n\n[/quote]
\n\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.
", "votes": 2, "replies": [] }, { "id": 92143, "author_name": "Sorceress", "author_url": "", "post_date": "2015-09-11T14:25:43.677000", "content": "@rcarson I'm not scared of this mistake, then why I have to hide it? A good point to learn something. It wasn't cheating. It was a mistake. And about private sharing, if it's needed, for sure, I'll explain exactly what happened to Admin and Admin then will decide.
", "votes": 1, "replies": [] }, { "id": 92114, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T12:29:16.750000", "content": "[quote=rcarson;92113]
\n\nThis feels wrong
\n\n[/quote]
\n\nWhat exactly feels wrong ?
", "votes": 1, "replies": [] }, { "id": 92115, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2015-09-11T12:32:31.943000", "content": "Two same score and a quick merge. Did you plan to merge after you two get the same score or before it?\nI am in no position to judge it but I doubt if you two achieve that score independently.
\n\nEdit: Basically I am against collaboration before merging.
", "votes": 2, "replies": [] }, { "id": 92173, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T17:17:57.180000", "content": "[quote=rcarson;92142]
\n\n@ DataGeek, I want to apologize if I sound offensive. I hope we could do as well as your team.
\n\nGood luck!
\n\n[/quote]
\n\nI think you were right at your place and was just trying to get the details. I am sorry for being rude.
", "votes": 1, "replies": [] }, { "id": 92121, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T12:52:52.780000", "content": "[quote=rcarson;92115]
\n\nEdit: Basically I am against collaboration before merging.
\n\n[/quote]
\n\nI have explained that pretty well and let me know if you still have any doubts.
", "votes": 1, "replies": [] }, { "id": 92202, "author_name": "Sudeep Juvekar", "author_url": "", "post_date": "2015-09-11T20:51:29.203000", "content": "Make it 3 new perfect scores....
", "votes": 2, "replies": [] }, { "id": 92177, "author_name": "omgponies", "author_url": "", "post_date": "2015-09-11T17:27:52.673000", "content": "\n\n\nAlthough the two perfect scores makes me think this could end up with Top 10 perfect scores
\n
I assumed it with the one score, but now as more come in very much think so, that there is likely some deterministic or leakage thing here, and then the end will be arrow splitting potentially off of the "other" model the leakage is mixed with (particularly if they redo the data and then we go back to the traditional models).
\n\nI saw something that I suspected may be what they are doing, but my code hasn't finished running yet, so not sure if it is real or apophenia at this point - will see this weekend I guess.
", "votes": 2, "replies": [] }, { "id": 92142, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2015-09-11T14:22:52.823000", "content": "@ DataGeek, I want to apologize if I sound offensive. I hope we could do as well as your team.
\n\nGood luck!
", "votes": 2, "replies": [] }, { "id": 92141, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2015-09-11T14:21:34.590000", "content": "[quote=DagnyT;92138]
\n\nKeyword - sometimes :D , it was a really very bad mistake, worth discussion? Or you just want us to be punished for it?
\n\n[/quote]
\n\nHey, I don't want you to be punished. I am just discussing what is 'private sharing' with DataGeek, and maybe everyone else.
\n\nFor me, it's Ok to make mistakes as long as people don't find out. : )
", "votes": 0, "replies": [] }, { "id": 92137, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2015-09-11T14:10:42.300000", "content": "[quote=DagnyT;92136]
\n\n@rcarson I really hope you don't make any mistake in your life, ideal one. Have a great day!
\n\n[/quote]
\n\nI make mistakes everyday and I am punished sometimes. : )
", "votes": 0, "replies": [] }, { "id": 92119, "author_name": "Chonglin Sun", "author_url": "", "post_date": "2015-09-11T12:45:28.550000", "content": "[quote=rcarson;92115]
\n\nTwo same score and a quick merge. Did you plan to merge after you two get the same score or before it?\nI am in no position to judge it but I doubt if you two achieve that score independently.
\n\nEdit: Basically I am against collaboration before merging.
\n\n[/quote]
\n\nTotally agree, if they didn't collaborate before merging, I don't think they have the necessity to collaborate at this time.
", "votes": 0, "replies": [] }, { "id": 92139, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T14:17:08.437000", "content": "[quote=rcarson;92132]
\n\n[quote=DataGeek;92126]
\n\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.
\n\n[/quote]
\n\nI am speaking my mind. I don't say you cheat. I don't mean it at all. \nBut the fact was that "Oleksii was testing the model (DagnyT's model?) with different seed and ended up submitting and getting the same score", when Oleksii and DagnyT were in two different teams. Yes, it was a mistake. But was it private sharing? When we talk about private sharing, does it have to do with just facts, or mistakes and intentions?
\n\nI'm no one and I don't think your mistakes need to be punished since it hurts no one. But if anyone "trying to make seen out of it", it's not me. Mistakes did get people removed, https://www.kaggle.com/c/seizure-prediction/forums/t/10972/the-one-with-only-21-entries-removed
\n\nI hope we all won't make such mistakes.
\n\n[/quote]
\n\nI agree with you and assure you there was no intention to cheat and no cheating has been done. We didn't try to use extra submission and you/admin can cross verify that with Kaggle that we only made 3 submission today. Also, admin can check all other previous submission of DagnyT and Our. It's our honesty that we explained what happened exactly. If admin finds we have done something wrong admins can remove us.
", "votes": 1, "replies": [] }, { "id": 92105, "author_name": "clustifier", "author_url": "", "post_date": "2015-09-11T11:35:37.887000", "content": "They probably merged afterwards.
\n\nI still have a window open with it.
\n\nHere is the screen capture.
", "votes": 1, "replies": [] }, { "id": 92113, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2015-09-11T12:26:20.803000", "content": "This feels wrong
", "votes": 0, "replies": [] }, { "id": 92120, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T12:45:36.533000", "content": "", "votes": -1, "replies": [] }, { "id": 92174, "author_name": "DataGeek", "author_url": "", "post_date": "2015-09-11T17:20:34.027000", "content": "[quote=rcarson;92164]
\n\n[quote=Little Boat;92160]
\n\nI would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)
\n\n[/quote]
\n\nHey, I am not the victim and this is just a thread that everybody earns some upvotes : D
\n\n[/quote]
\n\n+1
", "votes": 0, "replies": [] }, { "id": 92123, "author_name": "Jiwei Liu", "author_url": "", "post_date": "2015-09-11T12:58:59.927000", "content": "[quote=DataGeek;92121]
\n\n[quote=rcarson;92115]
\n\nEdit: Basically I am against collaboration before merging.
\n\n[/quote]
\n\nI have explained that pretty well and let me know if you still have any doubts.
\n\n[/quote]
\n\nThank you for explaining. What I feel doesn't matter, it is up to the admin to decide if this violates the 'no private sharing rule' or not.
", "votes": 0, "replies": [] }, { "id": 92094, "author_name": "Chonglin Sun", "author_url": "", "post_date": "2015-09-11T10:41:31.467000", "content": "I didn't see that, the merged together.
", "votes": 0, "replies": [] }, { "id": 92138, "author_name": "", "author_url": "", "post_date": "2015-09-11T14:13:22.687000", "content": "", "votes": -3, "replies": [] }, { "id": 92136, "author_name": "", "author_url": "", "post_date": "2015-09-11T14:09:23.090000", "content": "", "votes": -5, "replies": [] } ], "raw_markdown_by_id": { "92124": "Sounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!", "92175": "[quote=Bluefool;92124]\r\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!\r\n[/quote]\r\n\r\nA W E S O M E!!!, Thanks :D\r\n\r\nAlthough the two perfect scores makes me think this could end up with Top 10 perfect scores that won't include me XD. None of the those teams have [shared][1] any external data usage, so... there must be some magic features indeed.\r\n\r\nBack on topic, the Kaggle rules regarding private sharing are very clear, but its none of us job to enforce them, so just relax, if required, it will happen.\r\n\r\n\r\n [1]: https://www.kaggle.com/c/dato-native/forums/t/16272/external-data-repository", "92127": "[quote=Bluefool;92124]\r\n\r\nSounds like a case of bad timing so I wouldn't worry about it. I'm just rooting for NxGTR to get his Masters!\r\n\r\n[/quote]\r\n\r\nI don't need to mention that you are one of the few people who always say fair things :)", "92118": "[quote=rcarson;92115]\r\n\r\nTwo same score and a quick merge. Did you plan to merge after you two get the same score or before it?\r\nI am in no position to judge it but I doubt if you two achieve that score independently.\r\n\r\n[/quote]\r\n\r\nLet me clarify few things before people start running there minds in all direction\r\n\r\nDagnyT is the third member of DataMinders team (you can check past competition) and decided to start working on competition independently ( I don't know how long DagnyT started). Few hours back, DagnyT after submitting best model of current score shared the intention to form team and asked Oleksii to merge team. It just happen to be the case I was the Leader (Kaggle can confirm that) and Oleksii couldn't send the invite and I was busy in the office and got to know it later. They decided to merge repositories by the time I come back and form the team as soon as I see Oleksii message on hangout. It just happen to be the case Oleksii was testing the model with different seed and ended up submitting and getting the same score. Also Kaggle can confirm that we only made 3 submission in total including before and after the formation of team and best model was submitted today only. Also, I made Oleksii leader after seeing the message and then he formed the team.\r\n\r\nLet me make it clear what you were trying to say, DID YOU GUYS CHEAT? \r\n- answer is NO WE DIDN'T\r\n\r\nSorry If I am being rude but I just don't like people starting to run their mind without knowing what exactly happened.", "92160": "I would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)", "92164": "[quote=Little Boat;92160]\r\n\r\nI would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)\r\n\r\n[/quote]\r\n\r\nHey, I am not the victim and this is just a thread that everybody earns some upvotes : D", "92093": "That's strange!\r\n\r\n 212903,DagnyT,\"2015-09-11 07:19:46\",0.99805\r\n 207372,DataMinders,\"2015-09-11 09:57:55\",0.99805", "92185": "When I first saw the almost perfect score I thought there had to be something deterministic. I don't think there will be a perfect score though. If you look at very small files, there are some that are identical but one is a 1 and the other is a 0. At this point it would make sense to just patiently sift through some larger files and manually find small differences that are not captured by the standard features most people are using. For instance, I was looking for \"insignificant\" differences something as little as presence/absence of a space before an = sign in a tag, that might differentiate the underlying code generation method of native ads pages (or whatever StumbleUpon is calling native ads pages). Anyway, I really hope to be wrong, but this doesn't feel like a ML exercise at this point. ", "92132": "[quote=DataGeek;92126]\r\n\r\n\r\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.\r\n\r\n\r\n[/quote]\r\n\r\nI am speaking my mind. I don't say you cheat. I don't mean it at all. \r\nBut the fact was that \"Oleksii was testing the model (DagnyT's model?) with different seed and ended up submitting and getting the same score\", when Oleksii and DagnyT were in two different teams. Yes, it was a mistake. But was it private sharing? When we talk about private sharing, does it have to do with just facts, or mistakes and intentions?\r\n\r\nI'm no one and I don't think your mistakes need to be punished since it hurts no one. But if anyone \"trying to make seen out of it\", it's not me. Mistakes did get people removed, https://www.kaggle.com/c/seizure-prediction/forums/t/10972/the-one-with-only-21-entries-removed\r\n\r\nI hope we all won't make such mistakes.\r\n", "92126": "[quote=rcarson;92123]\r\n\r\n[quote=DataGeek;92121]\r\n\r\n[quote=rcarson;92115]\r\n\r\nEdit: Basically I am against collaboration before merging.\r\n\r\n[/quote]\r\n\r\nI have explained that pretty well and let me know if you still have any doubts.\r\n\r\n[/quote]\r\n\r\nThank you for explaining. What I feel doesn't matter, it is up to the admin to decide if this violates the 'no private sharing rule' or not.\r\n\r\n[/quote]\r\n\r\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.\r\n", "92143": "@rcarson I'm not scared of this mistake, then why I have to hide it? A good point to learn something. It wasn't cheating. It was a mistake. And about private sharing, if it's needed, for sure, I'll explain exactly what happened to Admin and Admin then will decide. ", "92114": "[quote=rcarson;92113]\r\n\r\nThis feels wrong\r\n\r\n[/quote]\r\n\r\nWhat exactly feels wrong ?", "92115": "Two same score and a quick merge. Did you plan to merge after you two get the same score or before it?\r\nI am in no position to judge it but I doubt if you two achieve that score independently.\r\n\r\nEdit: Basically I am against collaboration before merging.", "92173": "[quote=rcarson;92142]\r\n\r\n@ DataGeek, I want to apologize if I sound offensive. I hope we could do as well as your team. \r\n\r\nGood luck!\r\n\r\n[/quote]\r\n\r\nI think you were right at your place and was just trying to get the details. I am sorry for being rude.\r\n", "92121": "[quote=rcarson;92115]\r\n\r\nEdit: Basically I am against collaboration before merging.\r\n\r\n[/quote]\r\n\r\nI have explained that pretty well and let me know if you still have any doubts.", "92202": "Make it 3 new perfect scores....", "92177": "> Although the two perfect scores makes me think this could end up with Top 10 perfect scores\r\n\r\nI assumed it with the one score, but now as more come in very much think so, that there is likely some deterministic or leakage thing here, and then the end will be arrow splitting potentially off of the \"other\" model the leakage is mixed with (particularly if they redo the data and then we go back to the traditional models).\r\n\r\nI saw something that I suspected may be what they are doing, but my code hasn't finished running yet, so not sure if it is real or apophenia at this point - will see this weekend I guess.", "92142": "@ DataGeek, I want to apologize if I sound offensive. I hope we could do as well as your team. \r\n\r\nGood luck!", "92141": "[quote=DagnyT;92138]\r\n\r\nKeyword - sometimes :D , it was a really very bad mistake, worth discussion? Or you just want us to be punished for it? \r\n\r\n[/quote]\r\n\r\nHey, I don't want you to be punished. I am just discussing what is 'private sharing' with DataGeek, and maybe everyone else. \r\n\r\nFor me, it's Ok to make mistakes as long as people don't find out. : )", "92137": "[quote=DagnyT;92136]\r\n\r\n@rcarson I really hope you don't make any mistake in your life, ideal one. Have a great day! \r\n\r\n[/quote]\r\n\r\nI make mistakes everyday and I am punished sometimes. : )", "92119": "[quote=rcarson;92115]\r\n\r\nTwo same score and a quick merge. Did you plan to merge after you two get the same score or before it?\r\nI am in no position to judge it but I doubt if you two achieve that score independently.\r\n\r\nEdit: Basically I am against collaboration before merging.\r\n\r\n[/quote]\r\n\r\nTotally agree, if they didn't collaborate before merging, I don't think they have the necessity to collaborate at this time.", "92139": "[quote=rcarson;92132]\r\n\r\n[quote=DataGeek;92126]\r\n\r\n\r\nI don't think it violates anything but you are just trying to make seen out of it. I can guarantee anyone reading my explanation can understand it loud and clear. I frankly don't think anyone trying to cheat would make any such mistake. Only people playing fair can do such mistakes and you are right you really don't need to worry. I have never been caught cheating after participating in 61 competition because I don't do that and neither my team members will ever do that, they are hard working and fair. But I frankly didn't like your comment even after reading my explanation.\r\n\r\n\r\n[/quote]\r\n\r\nI am speaking my mind. I don't say you cheat. I don't mean it at all. \r\nBut the fact was that \"Oleksii was testing the model (DagnyT's model?) with different seed and ended up submitting and getting the same score\", when Oleksii and DagnyT were in two different teams. Yes, it was a mistake. But was it private sharing? When we talk about private sharing, does it have to do with just facts, or mistakes and intentions?\r\n\r\nI'm no one and I don't think your mistakes need to be punished since it hurts no one. But if anyone \"trying to make seen out of it\", it's not me. Mistakes did get people removed, https://www.kaggle.com/c/seizure-prediction/forums/t/10972/the-one-with-only-21-entries-removed\r\n\r\nI hope we all won't make such mistakes.\r\n\r\n\r\n[/quote]\r\n\r\nI agree with you and assure you there was no intention to cheat and no cheating has been done. We didn't try to use extra submission and you/admin can cross verify that with Kaggle that we only made 3 submission today. Also, admin can check all other previous submission of DagnyT and Our. It's our honesty that we explained what happened exactly. If admin finds we have done something wrong admins can remove us.\r\n", "92105": "They probably merged afterwards.\r\n\r\nI still have a window open with it.\r\n\r\nHere is the screen capture. ", "92113": "This feels wrong", "92120": "", "92174": "[quote=rcarson;92164]\r\n\r\n[quote=Little Boat;92160]\r\n\r\nI would say it is fine, nothing to worry. But you DID share before official merging, which people just pointed out and, of course, let it go. So I don't see why you guys think @rcarson is the bad guy here. IMO, he is the real victim in this post. :-)\r\n\r\n[/quote]\r\n\r\nHey, I am not the victim and this is just a thread that everybody earns some upvotes : D\r\n\r\n[/quote]\r\n\r\n+1", "92123": "[quote=DataGeek;92121]\r\n\r\n[quote=rcarson;92115]\r\n\r\nEdit: Basically I am against collaboration before merging.\r\n\r\n[/quote]\r\n\r\nI have explained that pretty well and let me know if you still have any doubts.\r\n\r\n[/quote]\r\n\r\nThank you for explaining. What I feel doesn't matter, it is up to the admin to decide if this violates the 'no private sharing rule' or not.", "92094": "I didn't see that, the merged together.", "92138": "", "92136": "" } }