If anyone look for teammate, let me know ...\nI just started by a LGBM model with different preprocessing !!
", "messages": [ { "id": 338383, "postDate": "2018-06-05T00:01:24.773Z", "content": "If anyone look for teammate, let me know ...\nI just started by a LGBM model with different preprocessing !!
", "rawMarkdown": "If anyone look for teammate, let me know ...\nI just started by a LGBM model with different preprocessing !!", "votes": 6 }, { "id": 338705, "postDate": "2018-06-05T16:13:34.357Z", "content": "WoW
", "rawMarkdown": "WoW", "votes": 1, "replies": [ { "id": 338909, "postDate": "2018-06-06T01:13:06.800Z", "content": "Hi my friend, I’m glad for you for the first medal, Stay focus YIANG :)
", "rawMarkdown": "Hi my friend, I’m glad for you for the first medal, Stay focus YIANG :)", "votes": 1 }, { "id": 338927, "postDate": "2018-06-06T01:54:12.853Z", "content": "Haha thanks, suffered a lot before that though, glad seeing you are back!
", "rawMarkdown": "Haha thanks, suffered a lot before that though, glad seeing you are back! ", "votes": 1 } ] }, { "id": 338583, "postDate": "2018-06-05T11:12:17.800Z", "content": "I'm still a beginner, but would love to join a team as well and contribute wherever I can! :)
", "rawMarkdown": "I'm still a beginner, but would love to join a team as well and contribute wherever I can! :)", "votes": 1, "replies": [ { "id": 338917, "postDate": "2018-06-06T01:27:44.120Z", "content": "Glad for you, but it seems your account just created 20 days ago
", "rawMarkdown": "Glad for you, but it seems your account just created 20 days ago ", "votes": 1 }, { "id": 338922, "postDate": "2018-06-06T01:36:27.447Z", "rawMarkdown": "", "votes": -1, "isDeleted": true } ] }, { "id": 338533, "postDate": "2018-06-05T09:12:44.970Z", "content": "I am a new kaggler here,get just 0.22 score by ridge and LGB,It will be so lucky to be in a team with you
", "rawMarkdown": "I am a new kaggler here,get just 0.22 score by ridge and LGB,It will be so lucky to be in a team with you ", "votes": 1, "replies": [ { "id": 338910, "postDate": "2018-06-06T01:15:30.137Z", "content": "Hi, Thank you for your reply !! I would glad to but i’m looking for someone who work with different models than LGBM as NN :)
", "rawMarkdown": "Hi, Thank you for your reply !! I would glad to but i’m looking for someone who work with different models than LGBM as NN :)", "votes": 1 }, { "id": 338971, "postDate": "2018-06-06T05:12:46.767Z", "content": "It' okay,thanks for your reply,I got only 0.2214 using .csv by LGB,what abou you?
", "rawMarkdown": "It' okay,thanks for your reply,I got only 0.2214 using .csv by LGB,what abou you?" }, { "id": 338972, "postDate": "2018-06-06T05:14:10.973Z", "rawMarkdown": "", "isDeleted": true }, { "id": 338973, "postDate": "2018-06-06T05:17:07.240Z", "content": "Great work, my lgb ; LB 0 2213
", "rawMarkdown": "Great work, my lgb ; LB 0 2213", "votes": 1 }, { "id": 338981, "postDate": "2018-06-06T05:23:31.300Z", "content": "wow,so the image data help alot?
", "rawMarkdown": "wow,so the image data help alot?" } ] }, { "id": 338409, "postDate": "2018-06-05T01:47:23.570Z", "content": "Hi~ Nice to meet you~ \nWe are interested. (Currently two people in the team.)\nI've sent you PM. Thanks~
", "rawMarkdown": "Hi~ Nice to meet you~ \nWe are interested. (Currently two people in the team.)\nI've sent you PM. Thanks~", "votes": 1, "replies": [ { "id": 341649, "postDate": "2018-06-12T02:56:47.220Z", "content": "Hi, are you still looking for teammate? I've sent you PM.
", "rawMarkdown": "Hi, are you still looking for teammate? I've sent you PM. " } ] }, { "id": 342050, "postDate": "2018-06-12T18:57:37.963Z", "content": "I am brand spanking new to this. I think that the more data you have the less likely you can figure out which way is up. I've got hunches that actually are based in reality. Awesome at calculus... but I NEVER lose sight of what's real. Part of the success that Warren Buffet attributes is that he lives AWAY from the NYSE. Called Etherium in Feb 2016. Would love to turn data science on its head. I'm truely gifted in finding what matters in this mess, in part because I'm a woman who has more connections in my brain. In part because I like math, yada, yada, yada... I can be off assistance in feeling the variables that matter.
", "rawMarkdown": "I am brand spanking new to this. I think that the more data you have the less likely you can figure out which way is up. I've got hunches that actually are based in reality. Awesome at calculus... but I NEVER lose sight of what's real. Part of the success that Warren Buffet attributes is that he lives AWAY from the NYSE. Called Etherium in Feb 2016. Would love to turn data science on its head. I'm truely gifted in finding what matters in this mess, in part because I'm a woman who has more connections in my brain. In part because I like math, yada, yada, yada... I can be off assistance in feeling the variables that matter.", "votes": -2 }, { "id": 342304, "postDate": "2018-06-13T09:24:35.447Z", "content": "Hi!\nI am looking for team-up :)
", "rawMarkdown": "Hi!\nI am looking for team-up :)" }, { "id": 340829, "postDate": "2018-06-10T12:32:10.453Z", "content": "Hello,
\n\nI have worked almost all my time building a set of solid features for the text-based data using deep learning & a language model. You won't find me on the leaderboard but what I can offer is sending my set of features for the training dataset, then you can see if you'd be interested in teaming up. I am just getting started with the final model using these features now there could be good synergies.
", "rawMarkdown": "Hello,\n\nI have worked almost all my time building a set of solid features for the text-based data using deep learning & a language model. You won't find me on the leaderboard but what I can offer is sending my set of features for the training dataset, then you can see if you'd be interested in teaming up. I am just getting started with the final model using these features now there could be good synergies. ", "replies": [ { "id": 340908, "postDate": "2018-06-10T15:53:37.073Z", "rawMarkdown": "", "votes": 1, "isDeleted": true } ] }, { "id": 339267, "postDate": "2018-06-06T16:21:00.143Z", "content": "Hi, me !
", "rawMarkdown": "Hi, me ! " }, { "id": 339064, "postDate": "2018-06-06T08:18:52.913Z", "content": "Hi Abdel, I would like to team up, but mine is a lightgbm too
", "rawMarkdown": "Hi Abdel, I would like to team up, but mine is a lightgbm too ", "replies": [ { "id": 339239, "postDate": "2018-06-06T15:09:41.897Z", "content": "Hi Abhimanyu, Yeah ...
", "rawMarkdown": "Hi Abhimanyu, Yeah ...", "votes": 1 } ] }, { "id": 338779, "postDate": "2018-06-05T18:25:01.753Z", "content": "Yes Abdel, count me in.
", "rawMarkdown": "Yes Abdel, count me in. ", "replies": [ { "id": 338907, "postDate": "2018-06-06T01:11:16.777Z", "content": "Sorry, but I don't see you on the leaderboard.
", "rawMarkdown": "Sorry, but I don't see you on the leaderboard.", "votes": 2 } ] }, { "id": 338571, "postDate": "2018-06-05T10:55:11.073Z", "content": "Hey there, I'm looking to team-up.
", "rawMarkdown": "Hey there, I'm looking to team-up.", "replies": [ { "id": 338911, "postDate": "2018-06-06T01:17:05.677Z", "content": "Sorry, but I don't see you on the leaderboard.
", "rawMarkdown": "Sorry, but I don't see you on the leaderboard.", "votes": 2 } ] }, { "id": 338389, "postDate": "2018-06-05T00:21:43.353Z", "content": "I'd be interested in teaming, although my local machine doesn't quite make the cut for this challenge. Lately I've been working with kernels, but would be open to suggestions if you had a preference for work environment.
\n\nEdit: I just realized how many entrants there are to this competition, and 53rd is higher up than I realized... no hard feelings if you're looking for someone more established than myself, I'm still learning.
", "rawMarkdown": "I'd be interested in teaming, although my local machine doesn't quite make the cut for this challenge. Lately I've been working with kernels, but would be open to suggestions if you had a preference for work environment. \n\nEdit: I just realized how many entrants there are to this competition, and 53rd is higher up than I realized... no hard feelings if you're looking for someone more established than myself, I'm still learning. ", "replies": [ { "id": 338390, "postDate": "2018-06-05T00:23:20.700Z", "content": "AWS works great if you're comfortable with the price. I've been using an r4.4xlarge for $1/hr. Otherwise, kernels should work.
", "rawMarkdown": "AWS works great if you're comfortable with the price. I've been using an r4.4xlarge for $1/hr. Otherwise, kernels should work.", "votes": 2 }, { "id": 338392, "postDate": "2018-06-05T00:32:15.417Z", "content": "I spent some time looking into renting something like AWS, but it seemed like something where I wouldn't really know how much I'm spending and it would sneak up on me... Would you mind telling me about how much you expect to spend on the whole competition? Things like time spent uploading data and idle time seem hard to account for.
\n\nNo worries if you're not comfortable disclosing that info, either way thank you for the suggestion.
", "rawMarkdown": "I spent some time looking into renting something like AWS, but it seemed like something where I wouldn't really know how much I'm spending and it would sneak up on me... Would you mind telling me about how much you expect to spend on the whole competition? Things like time spent uploading data and idle time seem hard to account for. \n\nNo worries if you're not comfortable disclosing that info, either way thank you for the suggestion. " }, { "id": 338393, "postDate": "2018-06-05T00:33:56.970Z", "content": "I'd guess I spend about $100-$200 on AWS throughout a typical competition. I can understand if this would be out of reach for some competitors and I'm excited that kernels are closing this gap.
", "rawMarkdown": "I'd guess I spend about $100-$200 on AWS throughout a typical competition. I can understand if this would be out of reach for some competitors and I'm excited that kernels are closing this gap.", "votes": 2 }, { "id": 338397, "postDate": "2018-06-05T00:44:11.413Z", "content": "Probably not the path I'll take, but thanks for sharing! Kernels are a huge help in that department, as well as helping newer kagglers get their feet on the ground, I've learned so much from public EDA/solution kernels.
", "rawMarkdown": "Probably not the path I'll take, but thanks for sharing! Kernels are a huge help in that department, as well as helping newer kagglers get their feet on the ground, I've learned so much from public EDA/solution kernels. " }, { "id": 338399, "postDate": "2018-06-05T00:47:27.107Z", "content": "Thanks @peterhurford for sharing details )
", "rawMarkdown": "Thanks @peterhurford for sharing details ) " }, { "id": 338926, "postDate": "2018-06-06T01:51:12.877Z", "content": "Sorry, but I don't see you on the leaderboard.
", "rawMarkdown": "Sorry, but I don't see you on the leaderboard." }, { "id": 339715, "postDate": "2018-06-07T13:27:03.253Z", "content": "I'm using AWS as well ! For this competition in particular I didn't need so much RAM. I'm using spot instance, for 32GB they are very cheap, like 0.06cts an hour and it can run longer than 6h :) For 64GB it might be like 30cts/h !
", "rawMarkdown": "I'm using AWS as well ! For this competition in particular I didn't need so much RAM. I'm using spot instance, for 32GB they are very cheap, like 0.06cts an hour and it can run longer than 6h :) For 64GB it might be like 30cts/h !\n", "votes": 1 }, { "id": 339757, "postDate": "2018-06-07T15:19:16.463Z", "content": "For someone starting late, could I expect to spend less than 50$ if I only used aws for the image processing and dealt with the csv locally?
", "rawMarkdown": "For someone starting late, could I expect to spend less than 50$ if I only used aws for the image processing and dealt with the csv locally?" }, { "id": 339763, "postDate": "2018-06-07T15:30:45.153Z", "content": "For someone starting late, don't use the images.
", "rawMarkdown": "For someone starting late, don't use the images." }, { "id": 339767, "postDate": "2018-06-07T15:44:25.553Z", "content": "Can you expand on that statement? I'm aware several people are performing well without the images, but I'd imagine the offer quite a bit of lift and are part of the challenge.
\n\nEven if my kaggle profile doesn't show much, I'm decently experienced with data like the csv, but have no hands on experience with images. I've followed along with the recent Google \nlandmark/fashion competitions and read a ton, but I've never really gotten into it. The barrier to entry is intimidating.
", "rawMarkdown": "Can you expand on that statement? I'm aware several people are performing well without the images, but I'd imagine the offer quite a bit of lift and are part of the challenge. \n\nEven if my kaggle profile doesn't show much, I'm decently experienced with data like the csv, but have no hands on experience with images. I've followed along with the recent Google \nlandmark/fashion competitions and read a ton, but I've never really gotten into it. The barrier to entry is intimidating. " }, { "id": 339774, "postDate": "2018-06-07T16:00:44.883Z", "content": "I think you just answered your own question. ;)
\n\nIf you are short on time prioritise the areas where you'll get the most bang for buck and in my view enough information is in the csv files to score well. This is particularly acute given as you say the barrier to entry is daunting and you have little experience with images - in that case simplify the problem down to what you are strong at. There are some good kernels which if digested fully and implemented well can be very useful.
\n\nYou can always come back to the images later if you have time or have run out of ideas but it isn't the obvious/easiest place to start.
", "rawMarkdown": "I think you just answered your own question. ;) \n\nIf you are short on time prioritise the areas where you'll get the most bang for buck and in my view enough information is in the csv files to score well. This is particularly acute given as you say the barrier to entry is daunting and you have little experience with images - in that case simplify the problem down to what you are strong at. There are some good kernels which if digested fully and implemented well can be very useful.\n\nYou can always come back to the images later if you have time or have run out of ideas but it isn't the obvious/easiest place to start.", "votes": 2 }, { "id": 339784, "postDate": "2018-06-07T16:28:45.553Z", "content": "Strictly from a placement point of view, that makes sense, and leaving images for last also makes sense, maybe there will come a competition soon where the time is right for me to dive into the images.
\n\nThanks for the feedback/direction.
", "rawMarkdown": "Strictly from a placement point of view, that makes sense, and leaving images for last also makes sense, maybe there will come a competition soon where the time is right for me to dive into the images.\n\nThanks for the feedback/direction. " } ] } ], "comments": [ { "id": 338705, "author_name": "YoungLamb", "author_url": "", "post_date": "2018-06-05T16:13:34.357000", "content": "WoW
", "votes": 1, "replies": [ { "id": 338909, "author_name": "Abdel (Casa)", "author_url": "", "post_date": "2018-06-06T01:13:06.800000", "content": "Hi my friend, I’m glad for you for the first medal, Stay focus YIANG :)
", "votes": 1, "replies": [] }, { "id": 338927, "author_name": "YoungLamb", "author_url": "", "post_date": "2018-06-06T01:54:12.853000", "content": "Haha thanks, suffered a lot before that though, glad seeing you are back!
", "votes": 1, "replies": [] } ] }, { "id": 338583, "author_name": "differentialmind", "author_url": "", "post_date": "2018-06-05T11:12:17.800000", "content": "I'm still a beginner, but would love to join a team as well and contribute wherever I can! :)
", "votes": 1, "replies": [ { "id": 338917, "author_name": "Abdel (Casa)", "author_url": "", "post_date": "2018-06-06T01:27:44.120000", "content": "Glad for you, but it seems your account just created 20 days ago
", "votes": 1, "replies": [] }, { "id": 338922, "author_name": "", "author_url": "", "post_date": "2018-06-06T01:36:27.447000", "content": "", "votes": -1, "replies": [] } ] }, { "id": 338533, "author_name": "ea", "author_url": "", "post_date": "2018-06-05T09:12:44.970000", "content": "I am a new kaggler here,get just 0.22 score by ridge and LGB,It will be so lucky to be in a team with you
", "votes": 1, "replies": [ { "id": 338910, "author_name": "Abdel (Casa)", "author_url": "", "post_date": "2018-06-06T01:15:30.137000", "content": "Hi, Thank you for your reply !! I would glad to but i’m looking for someone who work with different models than LGBM as NN :)
", "votes": 1, "replies": [] }, { "id": 338971, "author_name": "ea", "author_url": "", "post_date": "2018-06-06T05:12:46.767000", "content": "It' okay,thanks for your reply,I got only 0.2214 using .csv by LGB,what abou you?
", "votes": 0, "replies": [] }, { "id": 338972, "author_name": "", "author_url": "", "post_date": "2018-06-06T05:14:10.973000", "content": "", "votes": 0, "replies": [] }, { "id": 338973, "author_name": "Abdel (Casa)", "author_url": "", "post_date": "2018-06-06T05:17:07.240000", "content": "Great work, my lgb ; LB 0 2213
", "votes": 1, "replies": [] }, { "id": 338981, "author_name": "ea", "author_url": "", "post_date": "2018-06-06T05:23:31.300000", "content": "wow,so the image data help alot?
", "votes": 0, "replies": [] } ] }, { "id": 338409, "author_name": "khyeh", "author_url": "", "post_date": "2018-06-05T01:47:23.570000", "content": "Hi~ Nice to meet you~ \nWe are interested. (Currently two people in the team.)\nI've sent you PM. Thanks~
", "votes": 1, "replies": [ { "id": 341649, "author_name": "Xuan Cao", "author_url": "", "post_date": "2018-06-12T02:56:47.220000", "content": "Hi, are you still looking for teammate? I've sent you PM.
", "votes": 0, "replies": [] } ] }, { "id": 342050, "author_name": "Dark Indigo", "author_url": "", "post_date": "2018-06-12T18:57:37.963000", "content": "I am brand spanking new to this. I think that the more data you have the less likely you can figure out which way is up. I've got hunches that actually are based in reality. Awesome at calculus... but I NEVER lose sight of what's real. Part of the success that Warren Buffet attributes is that he lives AWAY from the NYSE. Called Etherium in Feb 2016. Would love to turn data science on its head. I'm truely gifted in finding what matters in this mess, in part because I'm a woman who has more connections in my brain. In part because I like math, yada, yada, yada... I can be off assistance in feeling the variables that matter.
", "votes": -2, "replies": [] }, { "id": 342304, "author_name": "AlexTru", "author_url": "", "post_date": "2018-06-13T09:24:35.447000", "content": "Hi!\nI am looking for team-up :)
", "votes": 0, "replies": [] }, { "id": 340829, "author_name": "Guillaume B", "author_url": "", "post_date": "2018-06-10T12:32:10.453000", "content": "Hello,
\n\nI have worked almost all my time building a set of solid features for the text-based data using deep learning & a language model. You won't find me on the leaderboard but what I can offer is sending my set of features for the training dataset, then you can see if you'd be interested in teaming up. I am just getting started with the final model using these features now there could be good synergies.
", "votes": 0, "replies": [ { "id": 340908, "author_name": "", "author_url": "", "post_date": "2018-06-10T15:53:37.073000", "content": "", "votes": 1, "replies": [] } ] }, { "id": 339267, "author_name": "__bathia_mkwr__", "author_url": "", "post_date": "2018-06-06T16:21:00.143000", "content": "Hi, me !
", "votes": 0, "replies": [] }, { "id": 339064, "author_name": "Abhimanyu Dikshit", "author_url": "", "post_date": "2018-06-06T08:18:52.913000", "content": "Hi Abdel, I would like to team up, but mine is a lightgbm too
", "votes": 0, "replies": [ { "id": 339239, "author_name": "Abdel (Casa)", "author_url": "", "post_date": "2018-06-06T15:09:41.897000", "content": "Hi Abhimanyu, Yeah ...
", "votes": 1, "replies": [] } ] }, { "id": 338779, "author_name": "Nagesh Somayajula", "author_url": "", "post_date": "2018-06-05T18:25:01.753000", "content": "Yes Abdel, count me in.
", "votes": 0, "replies": [ { "id": 338907, "author_name": "Abdel (Casa)", "author_url": "", "post_date": "2018-06-06T01:11:16.777000", "content": "Sorry, but I don't see you on the leaderboard.
", "votes": 2, "replies": [] } ] }, { "id": 338571, "author_name": "Temi Babs", "author_url": "", "post_date": "2018-06-05T10:55:11.073000", "content": "Hey there, I'm looking to team-up.
", "votes": 0, "replies": [ { "id": 338911, "author_name": "Abdel (Casa)", "author_url": "", "post_date": "2018-06-06T01:17:05.677000", "content": "Sorry, but I don't see you on the leaderboard.
", "votes": 2, "replies": [] } ] }, { "id": 338389, "author_name": "macfarll", "author_url": "", "post_date": "2018-06-05T00:21:43.353000", "content": "I'd be interested in teaming, although my local machine doesn't quite make the cut for this challenge. Lately I've been working with kernels, but would be open to suggestions if you had a preference for work environment.
\n\nEdit: I just realized how many entrants there are to this competition, and 53rd is higher up than I realized... no hard feelings if you're looking for someone more established than myself, I'm still learning.
", "votes": 0, "replies": [ { "id": 338390, "author_name": "Peter Hurford", "author_url": "", "post_date": "2018-06-05T00:23:20.700000", "content": "AWS works great if you're comfortable with the price. I've been using an r4.4xlarge for $1/hr. Otherwise, kernels should work.
", "votes": 2, "replies": [] }, { "id": 338392, "author_name": "macfarll", "author_url": "", "post_date": "2018-06-05T00:32:15.417000", "content": "I spent some time looking into renting something like AWS, but it seemed like something where I wouldn't really know how much I'm spending and it would sneak up on me... Would you mind telling me about how much you expect to spend on the whole competition? Things like time spent uploading data and idle time seem hard to account for.
\n\nNo worries if you're not comfortable disclosing that info, either way thank you for the suggestion.
", "votes": 0, "replies": [] }, { "id": 338393, "author_name": "Peter Hurford", "author_url": "", "post_date": "2018-06-05T00:33:56.970000", "content": "I'd guess I spend about $100-$200 on AWS throughout a typical competition. I can understand if this would be out of reach for some competitors and I'm excited that kernels are closing this gap.
", "votes": 2, "replies": [] }, { "id": 338397, "author_name": "macfarll", "author_url": "", "post_date": "2018-06-05T00:44:11.413000", "content": "Probably not the path I'll take, but thanks for sharing! Kernels are a huge help in that department, as well as helping newer kagglers get their feet on the ground, I've learned so much from public EDA/solution kernels.
", "votes": 0, "replies": [] }, { "id": 338399, "author_name": "Abdelwahed", "author_url": "", "post_date": "2018-06-05T00:47:27.107000", "content": "Thanks @peterhurford for sharing details )
", "votes": 0, "replies": [] }, { "id": 338926, "author_name": "Abdel (Casa)", "author_url": "", "post_date": "2018-06-06T01:51:12.877000", "content": "Sorry, but I don't see you on the leaderboard.
", "votes": 0, "replies": [] }, { "id": 339715, "author_name": "Antoine", "author_url": "", "post_date": "2018-06-07T13:27:03.253000", "content": "I'm using AWS as well ! For this competition in particular I didn't need so much RAM. I'm using spot instance, for 32GB they are very cheap, like 0.06cts an hour and it can run longer than 6h :) For 64GB it might be like 30cts/h !
", "votes": 1, "replies": [] }, { "id": 339757, "author_name": "macfarll", "author_url": "", "post_date": "2018-06-07T15:19:16.463000", "content": "For someone starting late, could I expect to spend less than 50$ if I only used aws for the image processing and dealt with the csv locally?
", "votes": 0, "replies": [] }, { "id": 339763, "author_name": "Mark Worrall", "author_url": "", "post_date": "2018-06-07T15:30:45.153000", "content": "For someone starting late, don't use the images.
", "votes": 0, "replies": [] }, { "id": 339767, "author_name": "macfarll", "author_url": "", "post_date": "2018-06-07T15:44:25.553000", "content": "Can you expand on that statement? I'm aware several people are performing well without the images, but I'd imagine the offer quite a bit of lift and are part of the challenge.
\n\nEven if my kaggle profile doesn't show much, I'm decently experienced with data like the csv, but have no hands on experience with images. I've followed along with the recent Google \nlandmark/fashion competitions and read a ton, but I've never really gotten into it. The barrier to entry is intimidating.
", "votes": 0, "replies": [] }, { "id": 339774, "author_name": "Mark Worrall", "author_url": "", "post_date": "2018-06-07T16:00:44.883000", "content": "I think you just answered your own question. ;)
\n\nIf you are short on time prioritise the areas where you'll get the most bang for buck and in my view enough information is in the csv files to score well. This is particularly acute given as you say the barrier to entry is daunting and you have little experience with images - in that case simplify the problem down to what you are strong at. There are some good kernels which if digested fully and implemented well can be very useful.
\n\nYou can always come back to the images later if you have time or have run out of ideas but it isn't the obvious/easiest place to start.
", "votes": 2, "replies": [] }, { "id": 339784, "author_name": "macfarll", "author_url": "", "post_date": "2018-06-07T16:28:45.553000", "content": "Strictly from a placement point of view, that makes sense, and leaving images for last also makes sense, maybe there will come a competition soon where the time is right for me to dive into the images.
\n\nThanks for the feedback/direction.
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "338383": "If anyone look for teammate, let me know ...\nI just started by a LGBM model with different preprocessing !!", "338705": "WoW", "338583": "I'm still a beginner, but would love to join a team as well and contribute wherever I can! :)", "338533": "I am a new kaggler here,get just 0.22 score by ridge and LGB,It will be so lucky to be in a team with you ", "338409": "Hi~ Nice to meet you~ \nWe are interested. (Currently two people in the team.)\nI've sent you PM. Thanks~", "342050": "I am brand spanking new to this. I think that the more data you have the less likely you can figure out which way is up. I've got hunches that actually are based in reality. Awesome at calculus... but I NEVER lose sight of what's real. Part of the success that Warren Buffet attributes is that he lives AWAY from the NYSE. Called Etherium in Feb 2016. Would love to turn data science on its head. I'm truely gifted in finding what matters in this mess, in part because I'm a woman who has more connections in my brain. In part because I like math, yada, yada, yada... I can be off assistance in feeling the variables that matter.", "342304": "Hi!\nI am looking for team-up :)", "340829": "Hello,\n\nI have worked almost all my time building a set of solid features for the text-based data using deep learning & a language model. You won't find me on the leaderboard but what I can offer is sending my set of features for the training dataset, then you can see if you'd be interested in teaming up. I am just getting started with the final model using these features now there could be good synergies. ", "339267": "Hi, me ! ", "339064": "Hi Abdel, I would like to team up, but mine is a lightgbm too ", "338779": "Yes Abdel, count me in. ", "338571": "Hey there, I'm looking to team-up.", "338389": "I'd be interested in teaming, although my local machine doesn't quite make the cut for this challenge. Lately I've been working with kernels, but would be open to suggestions if you had a preference for work environment. \n\nEdit: I just realized how many entrants there are to this competition, and 53rd is higher up than I realized... no hard feelings if you're looking for someone more established than myself, I'm still learning. " } }