Hi Kagglers. I am looking for a teammate to improve my 0.19456 position on the leaderboard. I am looking for 1-2 teammates who already have done submissions in this competition and scored at least 0.19.
\n\nAbout my submission:
\n\ntrain_numeric data (have not even looked at date and categorical information)About me
\n\nHi Kagglers. I am looking for a teammate to improve my 0.19456 position on the leaderboard. I am looking for 1-2 teammates who already have done submissions in this competition and scored at least 0.19.
\n\nAbout my submission:
\n\ntrain_numeric data (have not even looked at date and categorical information)About me
\n\nHello!
\n\nI live in SF area and would like to team up and we are a team of 2.\nPlease email to sb122109@gmail.com if you are interested.
\n\nThanks.
", "rawMarkdown": "Hello!\r\n\r\nI live in SF area and would like to team up and we are a team of 2.\r\nPlease email to sb122109@gmail.com if you are interested.\r\n\r\nThanks.", "votes": null }, { "id": "134461", "postDate": "09/06/2016 18:38:47", "content": "It does not look like you (or your team) have a submission on the leaderboard. This part who already have done submissions in this competition and scored at least 0.19 is one of the prerequisites.
", "rawMarkdown": "It does not look like you (or your team) have a submission on the leaderboard. This part **who already have done submissions in this competition and scored at least 0.19** is one of the prerequisites.", "votes": null }, { "id": "134720", "postDate": "09/08/2016 05:05:19", "content": "There is a Bay Area kaggle meetup in Burlingame on Mondays. I'm planning on going to the next one if you want to discuss. Otherwise feel free to send me a msg via LinkedIn (see my kaggle profile). I think we're in a similar spot. I've done a fairly simplistic submission using just the numeric data (and a dash of time) .
\n\nHere's the meetup info: https://www.meetup.com/StudyGroups/events/232658046/\nThere's usually a mix of complete novices with a few 'Kaggle Expert' rated people that show up (after 7pm).
", "rawMarkdown": "There is a Bay Area kaggle meetup in Burlingame on Mondays. I'm planning on going to the next one if you want to discuss. Otherwise feel free to send me a msg via LinkedIn (see my kaggle profile). I think we're in a similar spot. I've done a fairly simplistic submission using just the numeric data (and a dash of time) .\r\n\r\nHere's the meetup info: https://www.meetup.com/StudyGroups/events/232658046/\r\nThere's usually a mix of complete novices with a few 'Kaggle Expert' rated people that show up (after 7pm).", "votes": null }, { "id": "134755", "postDate": "09/08/2016 09:28:50", "content": "@SalvadorDali,
\n\nJust a remark. I you only accept people that have results as good as yours, then you're only getting people that have results exactly as good as yours ;) Indeed, why would a person having say a 0.20+ result accept to work with someone who has 'only' 0.19 when that person does not accept a 0.18?
\n\nI write this before I submit my first attempt so that I am not biased ;)
", "rawMarkdown": "SalvadorDali,\r\n\r\nJust a remark. I you only accept people that have results as good as yours, then you're only getting people that have results *exactly* as good as yours ;) Indeed, why would a person having say a 0.20+ result accept to work with someone who has 'only' 0.19 when that person does not accept a 0.18?\r\n\r\nI write this before I submit my first attempt so that I am not biased ;)", "votes": null }, { "id": "134758", "postDate": "09/08/2016 09:42:45", "content": "@CPMP currently there are 5 people who are above 0.19 and have results slightly below my score. Also there is a very very high chance I would have no problem of working with a person who has 0.18 or 0.17 score, but clearly not with someone who has score of 0 or nothing (mostly because this shows no commitment and most probably there will be no commitment in the future)
\n\nAs of why would a person with 0.2 score decide to work with me - there can be many many reasons. Will just give you a few:
\n\nThe list can go on, but I think that even these examples are enough to persuade you that sometimes it makes sense to join forces with someone who has reasonably worse score than you are.
\n\nIn my opinion if you have done your first solution after spending a day of thinking and will get 0.2, there is no point to create a team with a 0.19 guy (at least I would not do that). BTW, good luck with your first submission. Based on your record, it looks like you are kind of a guy who can score 0.22 from the first attempt
", "rawMarkdown": "CPMP currently there are 5 people who are above 0.19 and have results slightly below my score. Also there is a very very high chance I would have no problem of working with a person who has 0.18 or 0.17 score, but clearly not with someone who has score of 0 or nothing (mostly because this shows no commitment and most probably there will be no commitment in the future)\r\n\r\nAs of why would a person with 0.2 score decide to work with me - there can be many many reasons. Will just give you a few:\r\n\r\n - his classifier might require 2 days of running on a good machine, mine takes 5 hours on a bad machine\r\n - he might have used the whole dataset, mine uses only a part of it\r\n - he might have done hard feature engineering, I have not\r\n - he might have spent 2 weeks thinking and out of ideas, I barely started\r\n - he has done 50 submissions and most probably overfitting\r\n\r\nThe list can go on, but I think that even these examples are enough to persuade you that sometimes it makes sense to join forces with someone who has reasonably worse score than you are.\r\n\r\nIn my opinion if you have done your first solution after spending a day of thinking and will get 0.2, there is no point to create a team with a 0.19 guy (at least I would not do that). BTW, good luck with your first submission. Based on your record, it looks like you are kind of a guy who can score 0.22 from the first attempt", "votes": null }, { "id": "134797", "postDate": "09/08/2016 14:05:21", "content": "@CPMP\nI am also looking for people in the rank of 1-3 to team-up with me! x-)\nThe potential reason is that we could win the computation together.
", "rawMarkdown": "CPMP\r\nI am also looking for people in the rank of 1-3 to team-up with me! x-)\r\nThe potential reason is that we could win the computation together.", "votes": null }, { "id": "134799", "postDate": "09/08/2016 14:14:09", "content": "@SalvadorDali Hi,\nI also used only numeric features to get a score of 0.165. Would you like to share what did you use or we can have a chat via email?
", "rawMarkdown": "SalvadorDali Hi,\r\nI also used only numeric features to get a score of 0.165. Would you like to share what did you use or we can have a chat via email?", "votes": null } ], "comments": [ { "id": 134409, "author_name": "ss1231", "author_url": "", "post_date": "09/06/2016 14:45:22", "content": "Hello!
\n\nI live in SF area and would like to team up and we are a team of 2.\nPlease email to sb122109@gmail.com if you are interested.
\n\nThanks.
", "votes": null, "replies": [] }, { "id": 134461, "author_name": "salvadordali", "author_url": "", "post_date": "09/06/2016 18:38:47", "content": "It does not look like you (or your team) have a submission on the leaderboard. This part who already have done submissions in this competition and scored at least 0.19 is one of the prerequisites.
", "votes": null, "replies": [] }, { "id": 134720, "author_name": "datadave", "author_url": "", "post_date": "09/08/2016 05:05:19", "content": "There is a Bay Area kaggle meetup in Burlingame on Mondays. I'm planning on going to the next one if you want to discuss. Otherwise feel free to send me a msg via LinkedIn (see my kaggle profile). I think we're in a similar spot. I've done a fairly simplistic submission using just the numeric data (and a dash of time) .
\n\nHere's the meetup info: https://www.meetup.com/StudyGroups/events/232658046/\nThere's usually a mix of complete novices with a few 'Kaggle Expert' rated people that show up (after 7pm).
", "votes": null, "replies": [] }, { "id": 134755, "author_name": "cpmpml", "author_url": "", "post_date": "09/08/2016 09:28:50", "content": "@SalvadorDali,
\n\nJust a remark. I you only accept people that have results as good as yours, then you're only getting people that have results exactly as good as yours ;) Indeed, why would a person having say a 0.20+ result accept to work with someone who has 'only' 0.19 when that person does not accept a 0.18?
\n\nI write this before I submit my first attempt so that I am not biased ;)
", "votes": null, "replies": [] }, { "id": 134758, "author_name": "salvadordali", "author_url": "", "post_date": "09/08/2016 09:42:45", "content": "@CPMP currently there are 5 people who are above 0.19 and have results slightly below my score. Also there is a very very high chance I would have no problem of working with a person who has 0.18 or 0.17 score, but clearly not with someone who has score of 0 or nothing (mostly because this shows no commitment and most probably there will be no commitment in the future)
\n\nAs of why would a person with 0.2 score decide to work with me - there can be many many reasons. Will just give you a few:
\n\nThe list can go on, but I think that even these examples are enough to persuade you that sometimes it makes sense to join forces with someone who has reasonably worse score than you are.
\n\nIn my opinion if you have done your first solution after spending a day of thinking and will get 0.2, there is no point to create a team with a 0.19 guy (at least I would not do that). BTW, good luck with your first submission. Based on your record, it looks like you are kind of a guy who can score 0.22 from the first attempt
", "votes": null, "replies": [] }, { "id": 134797, "author_name": "mojtabaebrahimi85", "author_url": "", "post_date": "09/08/2016 14:05:21", "content": "@CPMP\nI am also looking for people in the rank of 1-3 to team-up with me! x-)\nThe potential reason is that we could win the computation together.
", "votes": null, "replies": [] }, { "id": 134799, "author_name": "zohairzohair4", "author_url": "", "post_date": "09/08/2016 14:14:09", "content": "@SalvadorDali Hi,\nI also used only numeric features to get a score of 0.165. Would you like to share what did you use or we can have a chat via email?
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "134283": "Hi Kagglers. I am looking for a teammate to improve my 0.19456 position on the leaderboard. I am looking for 1-2 teammates who already have done submissions in this competition and scored at least 0.19.\r\n\r\n**About my submission:**\r\n\r\n - 0.19456, 4 attempts\r\n - takes ~5 hours to generate on a crappy macbook i5 2.4ghz with 8gb ram\r\n - uses only `train_numeric` data (have not even looked at date and categorical information)\r\n - spent approximately 20 hours thinking about the solution\r\n\r\n**About me**\r\n\r\n- This is one of my first competitions and I am mostly looking for a person to exchange the knowledge and learn from each other\r\n- use python (pandas/scikit, other stuff), but have no problems working with an R guy\r\n- live in SF area, so if someone is from this area, we can meet. But location is not important at all.", "134409": "Hello!\r\n\r\nI live in SF area and would like to team up and we are a team of 2.\r\nPlease email to sb122109@gmail.com if you are interested.\r\n\r\nThanks.", "134461": "It does not look like you (or your team) have a submission on the leaderboard. This part **who already have done submissions in this competition and scored at least 0.19** is one of the prerequisites.", "134720": "There is a Bay Area kaggle meetup in Burlingame on Mondays. I'm planning on going to the next one if you want to discuss. Otherwise feel free to send me a msg via LinkedIn (see my kaggle profile). I think we're in a similar spot. I've done a fairly simplistic submission using just the numeric data (and a dash of time) .\r\n\r\nHere's the meetup info: https://www.meetup.com/StudyGroups/events/232658046/\r\nThere's usually a mix of complete novices with a few 'Kaggle Expert' rated people that show up (after 7pm).", "134755": "SalvadorDali,\r\n\r\nJust a remark. I you only accept people that have results as good as yours, then you're only getting people that have results *exactly* as good as yours ;) Indeed, why would a person having say a 0.20+ result accept to work with someone who has 'only' 0.19 when that person does not accept a 0.18?\r\n\r\nI write this before I submit my first attempt so that I am not biased ;)", "134758": "CPMP currently there are 5 people who are above 0.19 and have results slightly below my score. Also there is a very very high chance I would have no problem of working with a person who has 0.18 or 0.17 score, but clearly not with someone who has score of 0 or nothing (mostly because this shows no commitment and most probably there will be no commitment in the future)\r\n\r\nAs of why would a person with 0.2 score decide to work with me - there can be many many reasons. Will just give you a few:\r\n\r\n - his classifier might require 2 days of running on a good machine, mine takes 5 hours on a bad machine\r\n - he might have used the whole dataset, mine uses only a part of it\r\n - he might have done hard feature engineering, I have not\r\n - he might have spent 2 weeks thinking and out of ideas, I barely started\r\n - he has done 50 submissions and most probably overfitting\r\n\r\nThe list can go on, but I think that even these examples are enough to persuade you that sometimes it makes sense to join forces with someone who has reasonably worse score than you are.\r\n\r\nIn my opinion if you have done your first solution after spending a day of thinking and will get 0.2, there is no point to create a team with a 0.19 guy (at least I would not do that). BTW, good luck with your first submission. Based on your record, it looks like you are kind of a guy who can score 0.22 from the first attempt", "134797": "CPMP\r\nI am also looking for people in the rank of 1-3 to team-up with me! x-)\r\nThe potential reason is that we could win the computation together.", "134799": "SalvadorDali Hi,\r\nI also used only numeric features to get a score of 0.165. Would you like to share what did you use or we can have a chat via email?" }, "source": "meta" }