Congratulations to all the winners, and especially to Heng for his wonderful and selfless contributions and instructions. You continue to inspire and challenge us to do better all the time. A very well deserved first place.
\n\nThis has been a really fun and interesting competition. I have never done anything with speech data before, and this was a wonderful learning opportunity. Thank you Kaggle and Google for organizing this competition.
", "messages": [ { "id": 269565, "postDate": "2018-01-17T02:02:55.653Z", "content": "Congratulations to all the winners, and especially to Heng for his wonderful and selfless contributions and instructions. You continue to inspire and challenge us to do better all the time. A very well deserved first place.
\n\nThis has been a really fun and interesting competition. I have never done anything with speech data before, and this was a wonderful learning opportunity. Thank you Kaggle and Google for organizing this competition.
", "rawMarkdown": "Congratulations to all the winners, and especially to Heng for his wonderful and selfless contributions and instructions. You continue to inspire and challenge us to do better all the time. A very well deserved first place.\n\nThis has been a really fun and interesting competition. I have never done anything with speech data before, and this was a wonderful learning opportunity. Thank you Kaggle and Google for organizing this competition.", "votes": 23 }, { "id": 269647, "postDate": "2018-01-17T04:00:33.877Z", "content": "Thank you all. I hope everyone has a good time.
", "rawMarkdown": "Thank you all. I hope everyone has a good time.", "votes": 10, "replies": [ { "id": 269673, "postDate": "2018-01-17T05:02:11.477Z", "content": "Congrats CherKeng, Great to see you No 1. You gave so much.
", "rawMarkdown": "Congrats CherKeng, Great to see you No 1. You gave so much. ", "votes": 1 }, { "id": 269691, "postDate": "2018-01-17T05:40:40.927Z", "content": "@Heng, Congrats and thank you for sharing codes and ideas, very enjoy this competition
", "rawMarkdown": "@Heng, Congrats and thank you for sharing codes and ideas, very enjoy this competition " } ] }, { "id": 269668, "postDate": "2018-01-17T04:40:00.973Z", "content": "Heng, many congrats!!! Thank you much for sharing your work and ideas. This is very true Kaggle spirit, as I get it!
", "rawMarkdown": "Heng, many congrats!!! Thank you much for sharing your work and ideas. This is very true Kaggle spirit, as I get it!", "votes": 5 }, { "id": 269579, "postDate": "2018-01-17T02:27:09.287Z", "content": "Why is the number of teams less in the the Leaderboard Final Standings?
", "rawMarkdown": "Why is the number of teams less in the the Leaderboard Final Standings?", "votes": 3, "replies": [ { "id": 269585, "postDate": "2018-01-17T02:32:10.743Z", "content": "A lot of cheaters got removed.
", "rawMarkdown": "A lot of cheaters got removed.", "votes": 1 }, { "id": 269604, "postDate": "2018-01-17T02:51:24.247Z", "content": "Approximately 400 cheaters! 😮
", "rawMarkdown": "Approximately 400 cheaters! 😮", "votes": 2 }, { "id": 269620, "postDate": "2018-01-17T03:14:29.057Z", "rawMarkdown": "", "isDeleted": true }, { "id": 269639, "postDate": "2018-01-17T03:44:41.723Z", "content": "Unfortunately only about 20 users got removed that had better score than mine :)
", "rawMarkdown": "Unfortunately only about 20 users got removed that had better score than mine :)" }, { "id": 269657, "postDate": "2018-01-17T04:23:07.993Z", "content": "There are a couple of issues with the leaderboard that occurred when I switched the precision back up. They will be resolved shortly. Hang tight. (Only about 25 teams were removed from the LB.)
", "rawMarkdown": "There are a couple of issues with the leaderboard that occurred when I switched the precision back up. They will be resolved shortly. Hang tight. (Only about 25 teams were removed from the LB.)", "votes": 2 }, { "id": 269860, "postDate": "2018-01-17T12:16:40.083Z", "content": "How did they detect the cheaters? Distribution of right classifications?
", "rawMarkdown": "How did they detect the cheaters? Distribution of right classifications?" }, { "id": 269983, "postDate": "2018-01-17T16:04:20.453Z", "content": "@Edmundo Mizutani - If they tell you the algorithms all cheaters start to use it and just become more advanced cheaters
", "rawMarkdown": "@Edmundo Mizutani - If they tell you the algorithms all cheaters start to use it and just become more advanced cheaters", "votes": 1 }, { "id": 269990, "postDate": "2018-01-17T16:11:12.457Z", "content": "You are right, I would better restrain my curiosity :-)
", "rawMarkdown": "You are right, I would better restrain my curiosity :-)", "votes": 1 } ] }, { "id": 270209, "postDate": "2018-01-17T22:56:40.377Z", "content": "Congratulations to everyone, and thank you to everyone who helped out along the way, especially CherKeng. You really deserve it. Unfortunately I ran out of time to go through all the ideas I had. It was a tall order with my overwork on my full time job, being my first Kaggle and trying to read all the papers, learn all the frameworks as Python along the way. My undergrad course was in Comp Sci, AI & Psychology 20 years ago, and only recently jumped back on the wagon. I have been doing all of Andrew Ng's courses on Coursera and this was a great way to fill the gap till the last course in the specialization is released. I really found that the courses gave a good foundation, but the practical aspect of doing this challenge has been amazing and I did not expect the amazing level of support from the community and everyone trying to give people a leg up.
\n\nLooking forward to my next challenge.
", "rawMarkdown": "Congratulations to everyone, and thank you to everyone who helped out along the way, especially CherKeng. You really deserve it. Unfortunately I ran out of time to go through all the ideas I had. It was a tall order with my overwork on my full time job, being my first Kaggle and trying to read all the papers, learn all the frameworks as Python along the way. My undergrad course was in Comp Sci, AI & Psychology 20 years ago, and only recently jumped back on the wagon. I have been doing all of Andrew Ng's courses on Coursera and this was a great way to fill the gap till the last course in the specialization is released. I really found that the courses gave a good foundation, but the practical aspect of doing this challenge has been amazing and I did not expect the amazing level of support from the community and everyone trying to give people a leg up. \n\nLooking forward to my next challenge.", "votes": 1, "replies": [ { "id": 270222, "postDate": "2018-01-17T23:28:42.720Z", "content": "For a practical guide to applying deep learning, I highly recommend http://www.fast.ai/. Helped me enormously in this competition.
", "rawMarkdown": "For a practical guide to applying deep learning, I highly recommend http://www.fast.ai/. Helped me enormously in this competition." }, { "id": 270225, "postDate": "2018-01-17T23:33:00.673Z", "content": "Thanks, I did do the first section of fast.ai in between the other course. I quite like how fast.ai takes a top down approach and Andew Ng's courses take a bottom up. Hopefully meets somewhere in the middle :). Fast.ai got me started on Kaggle, I have done the Cats vs Dogs, but this is my first real competition.
", "rawMarkdown": "Thanks, I did do the first section of fast.ai in between the other course. I quite like how fast.ai takes a top down approach and Andew Ng's courses take a bottom up. Hopefully meets somewhere in the middle :). Fast.ai got me started on Kaggle, I have done the Cats vs Dogs, but this is my first real competition. " } ] }, { "id": 270146, "postDate": "2018-01-17T20:30:16.327Z", "content": "Heng,congrats!!! Thank you so much for sharing your solution,It helps a lot on my school project :D
", "rawMarkdown": "Heng,congrats!!! Thank you so much for sharing your solution,It helps a lot on my school project :D \n", "votes": 1 }, { "id": 269643, "postDate": "2018-01-17T03:55:34.820Z", "content": "To follow up Heng' code and idea, $500 credit is really no enough
", "rawMarkdown": "To follow up Heng' code and idea, $500 credit is really no enough", "votes": 1 }, { "id": 269638, "postDate": "2018-01-17T03:42:37.667Z", "content": "Congrats to Heng. This competition is a live learning session from the first winner.
", "rawMarkdown": "Congrats to Heng. This competition is a live learning session from the first winner.", "votes": 2 }, { "id": 270565, "postDate": "2018-01-18T14:24:42.230Z", "content": "Congrats
", "rawMarkdown": "Congrats" }, { "id": 270420, "postDate": "2018-01-18T07:40:25.623Z", "content": "Congratulations to everyone.
", "rawMarkdown": "Congratulations to everyone." }, { "id": 270140, "postDate": "2018-01-17T20:13:13.273Z", "content": "Congrats 😀
", "rawMarkdown": "Congrats 😀" }, { "id": 270050, "postDate": "2018-01-17T17:33:42.973Z", "content": "congrats - was a fun challenge. :)
", "rawMarkdown": "congrats - was a fun challenge. :)" }, { "id": 269628, "postDate": "2018-01-17T03:19:30.313Z", "rawMarkdown": "", "votes": 2, "isDeleted": true } ], "comments": [ { "id": 269647, "author_name": "hengck23", "author_url": "", "post_date": "2018-01-17T04:00:33.877000", "content": "Thank you all. I hope everyone has a good time.
", "votes": 10, "replies": [ { "id": 269673, "author_name": "Yibing Wu", "author_url": "", "post_date": "2018-01-17T05:02:11.477000", "content": "Congrats CherKeng, Great to see you No 1. You gave so much.
", "votes": 1, "replies": [] }, { "id": 269691, "author_name": "Hongta", "author_url": "", "post_date": "2018-01-17T05:40:40.927000", "content": "@Heng, Congrats and thank you for sharing codes and ideas, very enjoy this competition
", "votes": 0, "replies": [] } ] }, { "id": 269668, "author_name": "Alexander Popov", "author_url": "", "post_date": "2018-01-17T04:40:00.973000", "content": "Heng, many congrats!!! Thank you much for sharing your work and ideas. This is very true Kaggle spirit, as I get it!
", "votes": 5, "replies": [] }, { "id": 269579, "author_name": "SubhojeetPramanik", "author_url": "", "post_date": "2018-01-17T02:27:09.287000", "content": "Why is the number of teams less in the the Leaderboard Final Standings?
", "votes": 3, "replies": [ { "id": 269585, "author_name": "Bojan Tunguz", "author_url": "", "post_date": "2018-01-17T02:32:10.743000", "content": "A lot of cheaters got removed.
", "votes": 1, "replies": [] }, { "id": 269604, "author_name": "SubhojeetPramanik", "author_url": "", "post_date": "2018-01-17T02:51:24.247000", "content": "Approximately 400 cheaters! 😮
", "votes": 2, "replies": [] }, { "id": 269620, "author_name": "", "author_url": "", "post_date": "2018-01-17T03:14:29.057000", "content": "", "votes": 0, "replies": [] }, { "id": 269639, "author_name": "formigone", "author_url": "", "post_date": "2018-01-17T03:44:41.723000", "content": "Unfortunately only about 20 users got removed that had better score than mine :)
", "votes": 0, "replies": [] }, { "id": 269657, "author_name": "inversion", "author_url": "", "post_date": "2018-01-17T04:23:07.993000", "content": "There are a couple of issues with the leaderboard that occurred when I switched the precision back up. They will be resolved shortly. Hang tight. (Only about 25 teams were removed from the LB.)
", "votes": 2, "replies": [] }, { "id": 269860, "author_name": "Edmundo Mizutani", "author_url": "", "post_date": "2018-01-17T12:16:40.083000", "content": "How did they detect the cheaters? Distribution of right classifications?
", "votes": 0, "replies": [] }, { "id": 269983, "author_name": "Alex Shvets", "author_url": "", "post_date": "2018-01-17T16:04:20.453000", "content": "@Edmundo Mizutani - If they tell you the algorithms all cheaters start to use it and just become more advanced cheaters
", "votes": 1, "replies": [] }, { "id": 269990, "author_name": "Edmundo Mizutani", "author_url": "", "post_date": "2018-01-17T16:11:12.457000", "content": "You are right, I would better restrain my curiosity :-)
", "votes": 1, "replies": [] } ] }, { "id": 270209, "author_name": "Steve Bolton", "author_url": "", "post_date": "2018-01-17T22:56:40.377000", "content": "Congratulations to everyone, and thank you to everyone who helped out along the way, especially CherKeng. You really deserve it. Unfortunately I ran out of time to go through all the ideas I had. It was a tall order with my overwork on my full time job, being my first Kaggle and trying to read all the papers, learn all the frameworks as Python along the way. My undergrad course was in Comp Sci, AI & Psychology 20 years ago, and only recently jumped back on the wagon. I have been doing all of Andrew Ng's courses on Coursera and this was a great way to fill the gap till the last course in the specialization is released. I really found that the courses gave a good foundation, but the practical aspect of doing this challenge has been amazing and I did not expect the amazing level of support from the community and everyone trying to give people a leg up.
\n\nLooking forward to my next challenge.
", "votes": 1, "replies": [ { "id": 270222, "author_name": "bsp2020", "author_url": "", "post_date": "2018-01-17T23:28:42.720000", "content": "For a practical guide to applying deep learning, I highly recommend http://www.fast.ai/. Helped me enormously in this competition.
", "votes": 0, "replies": [] }, { "id": 270225, "author_name": "Steve Bolton", "author_url": "", "post_date": "2018-01-17T23:33:00.673000", "content": "Thanks, I did do the first section of fast.ai in between the other course. I quite like how fast.ai takes a top down approach and Andew Ng's courses take a bottom up. Hopefully meets somewhere in the middle :). Fast.ai got me started on Kaggle, I have done the Cats vs Dogs, but this is my first real competition.
", "votes": 0, "replies": [] } ] }, { "id": 270146, "author_name": "Meds", "author_url": "", "post_date": "2018-01-17T20:30:16.327000", "content": "Heng,congrats!!! Thank you so much for sharing your solution,It helps a lot on my school project :D
", "votes": 1, "replies": [] }, { "id": 269643, "author_name": "Hongta", "author_url": "", "post_date": "2018-01-17T03:55:34.820000", "content": "To follow up Heng' code and idea, $500 credit is really no enough
", "votes": 1, "replies": [] }, { "id": 269638, "author_name": "Josan", "author_url": "", "post_date": "2018-01-17T03:42:37.667000", "content": "Congrats to Heng. This competition is a live learning session from the first winner.
", "votes": 2, "replies": [] }, { "id": 270565, "author_name": "Meow Agent", "author_url": "", "post_date": "2018-01-18T14:24:42.230000", "content": "Congrats
", "votes": 0, "replies": [] }, { "id": 270420, "author_name": "MyatThuHein", "author_url": "", "post_date": "2018-01-18T07:40:25.623000", "content": "Congratulations to everyone.
", "votes": 0, "replies": [] }, { "id": 270140, "author_name": "Gaurav Kumar", "author_url": "", "post_date": "2018-01-17T20:13:13.273000", "content": "Congrats 😀
", "votes": 0, "replies": [] }, { "id": 270050, "author_name": "Johannes Shr", "author_url": "", "post_date": "2018-01-17T17:33:42.973000", "content": "congrats - was a fun challenge. :)
", "votes": 0, "replies": [] }, { "id": 269628, "author_name": "", "author_url": "", "post_date": "2018-01-17T03:19:30.313000", "content": "", "votes": 2, "replies": [] } ], "raw_markdown_by_id": { "269565": "Congratulations to all the winners, and especially to Heng for his wonderful and selfless contributions and instructions. You continue to inspire and challenge us to do better all the time. A very well deserved first place.\n\nThis has been a really fun and interesting competition. I have never done anything with speech data before, and this was a wonderful learning opportunity. Thank you Kaggle and Google for organizing this competition.", "269647": "Thank you all. I hope everyone has a good time.", "269668": "Heng, many congrats!!! Thank you much for sharing your work and ideas. This is very true Kaggle spirit, as I get it!", "269579": "Why is the number of teams less in the the Leaderboard Final Standings?", "270209": "Congratulations to everyone, and thank you to everyone who helped out along the way, especially CherKeng. You really deserve it. Unfortunately I ran out of time to go through all the ideas I had. It was a tall order with my overwork on my full time job, being my first Kaggle and trying to read all the papers, learn all the frameworks as Python along the way. My undergrad course was in Comp Sci, AI & Psychology 20 years ago, and only recently jumped back on the wagon. I have been doing all of Andrew Ng's courses on Coursera and this was a great way to fill the gap till the last course in the specialization is released. I really found that the courses gave a good foundation, but the practical aspect of doing this challenge has been amazing and I did not expect the amazing level of support from the community and everyone trying to give people a leg up. \n\nLooking forward to my next challenge.", "270146": "Heng,congrats!!! Thank you so much for sharing your solution,It helps a lot on my school project :D \n", "269643": "To follow up Heng' code and idea, $500 credit is really no enough", "269638": "Congrats to Heng. This competition is a live learning session from the first winner.", "270565": "Congrats", "270420": "Congratulations to everyone.", "270140": "Congrats 😀", "270050": "congrats - was a fun challenge. :)", "269628": "" } }