{ "id": 45716, "title": "Congrats to new kaggle No.1", "url": "/competitions/cdiscount-image-classification-challenge/discussion/45716", "author_name": "", "post_date": "2017-12-15T00:36:36.947288800Z", "votes": 38, "comment_count": 19, "views": 0, "content": "

It takes a lot of dedication and effort to rise to the top spot.

\n\n

It was enjoyable watching you rising gradually through ranks with your awesome DL skills.

\n\n

Well done on achieving it (and in less than 2 years) .

\n\n

Edit: Also congrats to the winners and generally all people in the top 10. This was a very demanding competition

", "messages": [ { "id": "257803", "postDate": "12/15/2017 00:36:36", "content": "

It takes a lot of dedication and effort to rise to the top spot.

\n\n

It was enjoyable watching you rising gradually through ranks with your awesome DL skills.

\n\n

Well done on achieving it (and in less than 2 years) .

\n\n

Edit: Also congrats to the winners and generally all people in the top 10. This was a very demanding competition

", "rawMarkdown": "It takes a lot of dedication and effort to rise to the top spot. \n\nIt was enjoyable watching you rising gradually through ranks with your awesome DL skills. \n\nWell done on achieving it (and in less than 2 years) . \n\n**Edit**: Also congrats to the winners and generally all people in the top 10. This was a very demanding competition", "votes": null }, { "id": "257806", "postDate": "12/15/2017 00:38:26", "content": "

bestfitting is indeed best fitting! Congrats ! Hope to have a chance to work with you! Interested to know your error and data analysis.

", "rawMarkdown": "bestfitting is indeed best fitting! Congrats ! Hope to have a chance to work with you! Interested to know your error and data analysis.", "votes": null }, { "id": "257813", "postDate": "12/15/2017 00:47:14", "content": "

Congratulations @besfitting. That is quite an achievement kudos.

", "rawMarkdown": "Congratulations @besfitting. That is quite an achievement kudos.", "votes": null }, { "id": "257875", "postDate": "12/15/2017 02:43:32", "content": "

Great thanks to you,I can not imagine this when I started kaggle competition 15 months ago,so lucky!\n
Kaggle is great platform,so many great datasets,solutions and discussions here.I keep reading solutions from you and other kagglers , reading papers related and taking part in competitions to let me concentrated on learning and evaluating some algorithms.
\n
The kaggle rank is an honour,the experiences are treasures,and the community is unique!\n
Thanks to all of you!

", "rawMarkdown": "Great thanks to you,I can not imagine this when I started kaggle competition 15 months ago,so lucky!\n
Kaggle is great platform,so many great datasets,solutions and discussions here.I keep reading solutions from you and other kagglers , reading papers related and taking part in competitions to let me concentrated on learning and evaluating some algorithms.
\n
The kaggle rank is an honour,the experiences are treasures,and the community is unique!\n
Thanks to all of you!", "votes": null }, { "id": "257876", "postDate": "12/15/2017 02:44:27", "content": "

Thanks!

", "rawMarkdown": "Thanks!", "votes": null }, { "id": "257888", "postDate": "12/15/2017 03:00:35", "content": "

You are always generous and warmhearted,you make kaggle platform greater.it will be very interesting to work with you!
\nAs to error and data analysis,I mainly start from data distribution,and then I will evaluate the prediction distribution,the change distribution between two predictions,and visualization,and sometimes I will do some kind of simulation.This is very important in real project and a competition.So I think its very important to make use of statistics and probability and related tools.

", "rawMarkdown": "You are always generous and warmhearted,you make kaggle platform greater.it will be very interesting to work with you!
\nAs to error and data analysis,I mainly start from data distribution,and then I will evaluate the prediction distribution,the change distribution between two predictions,and visualization,and sometimes I will do some kind of simulation.This is very important in real project and a competition.So I think its very important to make use of statistics and probability and related tools.", "votes": null }, { "id": "257896", "postDate": "12/15/2017 03:36:15", "content": "

Congrats @bestfitting

", "rawMarkdown": "Congrats @bestfitting", "votes": null }, { "id": "257897", "postDate": "12/15/2017 03:37:13", "content": "

Congrats, @bestfitting!

", "rawMarkdown": "Congrats, @bestfitting!", "votes": null }, { "id": "257900", "postDate": "12/15/2017 03:52:09", "content": "

Our team is doing the exact same thing last night! So you will notice that we are trying to catch up last night. We are comparing submitted distribution and trying to figure out what is the best way to ensemble. But we miss out the part on simulation.

\n\n

We will post our findings after some organisation. Hope to exchange some insights on that!

", "rawMarkdown": "Our team is doing the exact same thing last night! So you will notice that we are trying to catch up last night. We are comparing submitted distribution and trying to figure out what is the best way to ensemble. But we miss out the part on simulation.\n\nWe will post our findings after some organisation. Hope to exchange some insights on that!", "votes": null }, { "id": "258032", "postDate": "12/15/2017 09:29:38", "content": "

Congratulations! Acing all those (and not only) CV competitions is quite a feat!

", "rawMarkdown": "Congratulations! Acing all those (and not only) CV competitions is quite a feat!", "votes": null }, { "id": "258042", "postDate": "12/15/2017 10:18:35", "content": "

congrats friend... good work keep it up

", "rawMarkdown": "congrats friend... good work keep it up", "votes": null }, { "id": "258074", "postDate": "12/15/2017 11:51:06", "content": "

Congrats! A very well deserved victory!

", "rawMarkdown": "Congrats! A very well deserved victory!", "votes": null }, { "id": "258140", "postDate": "12/15/2017 15:39:31", "content": "

Congrats @bestfitting

", "rawMarkdown": "Congrats @bestfitting", "votes": null }, { "id": "258242", "postDate": "12/15/2017 19:52:05", "content": "

You should share more of your wisdom! It is quite interesting what approaches did you use and even better to get some code :) Sharing is the most efficient way for us as community to improve.

\n\n

P. S. Still waiting details on your amazing performance in other competitions too!

", "rawMarkdown": "You should share more of your wisdom! It is quite interesting what approaches did you use and even better to get some code :) Sharing is the most efficient way for us as community to improve. \n\nP. S. Still waiting details on your amazing performance in other competitions too!", "votes": null }, { "id": "258266", "postDate": "12/15/2017 20:59:02", "content": "
\n

You should share more of your wisdom! It is quite interesting what approaches did you use and even better to get some code :) Sharing is the most efficient way for us as community to improve.

\n
\n\n

He said he would write a summary within 2 days in this discussion

", "rawMarkdown": ">You should share more of your wisdom! It is quite interesting what approaches did you use and even better to get some code :) Sharing is the most efficient way for us as community to improve.\n\nHe said he would write a summary within 2 days in this [discussion][1] \n\n\n [1]: https://www.kaggle.com/c/cdiscount-image-classification-challenge/discussion/45711", "votes": null }, { "id": "258688", "postDate": "12/16/2017 19:43:30", "content": "

Thanks to @Shujian.M.Coder,＠ShubhajitML,＠Wojtek Rosinski,@Jay,@Tantali,@arigarcia,@BuryBuryZymon,@Ashish Chopra

", "rawMarkdown": "Thanks to @Shujian.M.Coder,＠ShubhajitML,＠Wojtek Rosinski,@Jay,@Tantali,@arigarcia,@BuryBuryZymon,@Ashish Chopra", "votes": null }, { "id": "259598", "postDate": "12/18/2017 17:30:53", "content": "

Congratulations @bestfitting :)

", "rawMarkdown": "Congratulations @bestfitting :)", "votes": null }, { "id": "259919", "postDate": "12/19/2017 07:21:01", "content": "

Congrats @bestfitting!

", "rawMarkdown": "Congrats @bestfitting!", "votes": null }, { "id": "260381", "postDate": "12/20/2017 05:08:08", "content": "

what tools do you use to data distribution analysis? thanks

", "rawMarkdown": "what tools do you use to data distribution analysis? thanks", "votes": null }, { "id": "260441", "postDate": "12/20/2017 07:46:53", "content": "

I like the tools such as:scipy.stats and seaborn,the best way to get familiar with them is to take part in a machine learning competition and try to learn from others.A lot of kagglers are very very strong,we can read the posts and kernels from them.

", "rawMarkdown": "I like the tools such as:scipy.stats and seaborn,the best way to get familiar with them is to take part in a machine learning competition and try to learn from others.A lot of kagglers are very very strong,we can read the posts and kernels from them.", "votes": null } ], "comments": [ { "id": 257806, "author_name": "hengck23", "author_url": "", "post_date": "12/15/2017 00:38:26", "content": "

bestfitting is indeed best fitting! Congrats ! Hope to have a chance to work with you! Interested to know your error and data analysis.

", "votes": null, "replies": [ { "id": 257888, "author_name": "bestfitting", "author_url": "", "post_date": "12/15/2017 03:00:35", "content": "

You are always generous and warmhearted,you make kaggle platform greater.it will be very interesting to work with you!
\nAs to error and data analysis,I mainly start from data distribution,and then I will evaluate the prediction distribution,the change distribution between two predictions,and visualization,and sometimes I will do some kind of simulation.This is very important in real project and a competition.So I think its very important to make use of statistics and probability and related tools.

", "votes": null, "replies": [] }, { "id": 257900, "author_name": "hengck23", "author_url": "", "post_date": "12/15/2017 03:52:09", "content": "

Our team is doing the exact same thing last night! So you will notice that we are trying to catch up last night. We are comparing submitted distribution and trying to figure out what is the best way to ensemble. But we miss out the part on simulation.

\n\n

We will post our findings after some organisation. Hope to exchange some insights on that!

", "votes": null, "replies": [] }, { "id": 258242, "author_name": "ceperaang", "author_url": "", "post_date": "12/15/2017 19:52:05", "content": "

You should share more of your wisdom! It is quite interesting what approaches did you use and even better to get some code :) Sharing is the most efficient way for us as community to improve.

\n\n

P. S. Still waiting details on your amazing performance in other competitions too!

", "votes": null, "replies": [] }, { "id": 258266, "author_name": "serigne", "author_url": "", "post_date": "12/15/2017 20:59:02", "content": "
\n

You should share more of your wisdom! It is quite interesting what approaches did you use and even better to get some code :) Sharing is the most efficient way for us as community to improve.

\n
\n\n

He said he would write a summary within 2 days in this discussion

", "votes": null, "replies": [] }, { "id": 260381, "author_name": "juguangliang", "author_url": "", "post_date": "12/20/2017 05:08:08", "content": "

what tools do you use to data distribution analysis? thanks

", "votes": null, "replies": [] }, { "id": 260441, "author_name": "bestfitting", "author_url": "", "post_date": "12/20/2017 07:46:53", "content": "

I like the tools such as:scipy.stats and seaborn,the best way to get familiar with them is to take part in a machine learning competition and try to learn from others.A lot of kagglers are very very strong,we can read the posts and kernels from them.

", "votes": null, "replies": [] } ] }, { "id": 257813, "author_name": "sheriytm", "author_url": "", "post_date": "12/15/2017 00:47:14", "content": "

Congratulations @besfitting. That is quite an achievement kudos.

", "votes": null, "replies": [ { "id": 257876, "author_name": "bestfitting", "author_url": "", "post_date": "12/15/2017 02:44:27", "content": "

Thanks!

", "votes": null, "replies": [] } ] }, { "id": 257875, "author_name": "bestfitting", "author_url": "", "post_date": "12/15/2017 02:43:32", "content": "

Great thanks to you,I can not imagine this when I started kaggle competition 15 months ago,so lucky!\n
Kaggle is great platform,so many great datasets,solutions and discussions here.I keep reading solutions from you and other kagglers , reading papers related and taking part in competitions to let me concentrated on learning and evaluating some algorithms.
\n
The kaggle rank is an honour,the experiences are treasures,and the community is unique!\n
Thanks to all of you!

", "votes": null, "replies": [ { "id": 257897, "author_name": "shujian", "author_url": "", "post_date": "12/15/2017 03:37:13", "content": "

Congrats, @bestfitting!

", "votes": null, "replies": [] }, { "id": 259919, "author_name": "ashishchopra", "author_url": "", "post_date": "12/19/2017 07:21:01", "content": "

Congrats @bestfitting!

", "votes": null, "replies": [] } ] }, { "id": 257896, "author_name": "shubhajitdas", "author_url": "", "post_date": "12/15/2017 03:36:15", "content": "

Congrats @bestfitting

", "votes": null, "replies": [] }, { "id": 258032, "author_name": "wrosinski", "author_url": "", "post_date": "12/15/2017 09:29:38", "content": "

Congratulations! Acing all those (and not only) CV competitions is quite a feat!

", "votes": null, "replies": [] }, { "id": 258042, "author_name": "jssolanke213", "author_url": "", "post_date": "12/15/2017 10:18:35", "content": "

congrats friend... good work keep it up

", "votes": null, "replies": [] }, { "id": 258074, "author_name": "tantali", "author_url": "", "post_date": "12/15/2017 11:51:06", "content": "

Congrats! A very well deserved victory!

", "votes": null, "replies": [] }, { "id": 258140, "author_name": "aristofanio", "author_url": "", "post_date": "12/15/2017 15:39:31", "content": "

Congrats @bestfitting

", "votes": null, "replies": [] }, { "id": 258688, "author_name": "bestfitting", "author_url": "", "post_date": "12/16/2017 19:43:30", "content": "

Thanks to @Shujian.M.Coder,＠ShubhajitML,＠Wojtek Rosinski,@Jay,@Tantali,@arigarcia,@BuryBuryZymon,@Ashish Chopra

", "votes": null, "replies": [] }, { "id": 259598, "author_name": "maheshdadhich", "author_url": "", "post_date": "12/18/2017 17:30:53", "content": "

Congratulations @bestfitting :)

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "257803": "It takes a lot of dedication and effort to rise to the top spot. \n\nIt was enjoyable watching you rising gradually through ranks with your awesome DL skills. \n\nWell done on achieving it (and in less than 2 years) . \n\n**Edit**: Also congrats to the winners and generally all people in the top 10. This was a very demanding competition", "257806": "bestfitting is indeed best fitting! Congrats ! Hope to have a chance to work with you! Interested to know your error and data analysis.", "257813": "Congratulations @besfitting. That is quite an achievement kudos.", "257875": "Great thanks to you,I can not imagine this when I started kaggle competition 15 months ago,so lucky!\n
Kaggle is great platform,so many great datasets,solutions and discussions here.I keep reading solutions from you and other kagglers , reading papers related and taking part in competitions to let me concentrated on learning and evaluating some algorithms.
\n
The kaggle rank is an honour,the experiences are treasures,and the community is unique!\n
Thanks to all of you!", "257876": "Thanks!", "257888": "You are always generous and warmhearted,you make kaggle platform greater.it will be very interesting to work with you!
\nAs to error and data analysis,I mainly start from data distribution,and then I will evaluate the prediction distribution,the change distribution between two predictions,and visualization,and sometimes I will do some kind of simulation.This is very important in real project and a competition.So I think its very important to make use of statistics and probability and related tools.", "257896": "Congrats @bestfitting", "257897": "Congrats, @bestfitting!", "257900": "Our team is doing the exact same thing last night! So you will notice that we are trying to catch up last night. We are comparing submitted distribution and trying to figure out what is the best way to ensemble. But we miss out the part on simulation.\n\nWe will post our findings after some organisation. Hope to exchange some insights on that!", "258032": "Congratulations! Acing all those (and not only) CV competitions is quite a feat!", "258042": "congrats friend... good work keep it up", "258074": "Congrats! A very well deserved victory!", "258140": "Congrats @bestfitting", "258242": "You should share more of your wisdom! It is quite interesting what approaches did you use and even better to get some code :) Sharing is the most efficient way for us as community to improve. \n\nP. S. Still waiting details on your amazing performance in other competitions too!", "258266": ">You should share more of your wisdom! It is quite interesting what approaches did you use and even better to get some code :) Sharing is the most efficient way for us as community to improve.\n\nHe said he would write a summary within 2 days in this [discussion][1] \n\n\n [1]: https://www.kaggle.com/c/cdiscount-image-classification-challenge/discussion/45711", "258688": "Thanks to @Shujian.M.Coder,＠ShubhajitML,＠Wojtek Rosinski,@Jay,@Tantali,@arigarcia,@BuryBuryZymon,@Ashish Chopra", "259598": "Congratulations @bestfitting :)", "259919": "Congrats @bestfitting!", "260381": "what tools do you use to data distribution analysis? thanks", "260441": "I like the tools such as:scipy.stats and seaborn,the best way to get familiar with them is to take part in a machine learning competition and try to learn from others.A lot of kagglers are very very strong,we can read the posts and kernels from them." }, "source": "meta" }