![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F13477216%2F3783731c7eb4fcbba2d5c5a95b52a439%2Fhomer.png?generation=1677526274761602&alt=media\")
For me, I'm ready for some shake-down cause the submissions of this week are all about overfitting the LB.
", "messages": [ { "id": 2161197, "postDate": "2023-02-27T10:05:37.693Z", "content": "For me, I'm ready for some shake-down cause the submissions of this week are all about overfitting the LB.
", "rawMarkdown": "For me, I'm ready for some shake-down cause the submissions of this week are all about overfitting the LB.", "votes": 22 }, { "id": 2161556, "postDate": "2023-02-27T15:21:06.497Z", "content": "as a new father, the shakeup reminds me of a song: \"the people on the bus go up and down, up and down, up and down…\"
\nbest wishes to all the ones that put in the effort in the game.. and good night(for those in GMT+8)
", "rawMarkdown": "as a new father, the shakeup reminds me of a song: \"the people on the bus go up and down, up and down, up and down...\"\n\nbest wishes to all the ones that put in the effort in the game.. and good night(for those in GMT+8)", "votes": 14, "replies": [ { "id": 2161849, "postDate": "2023-02-27T19:31:42.380Z", "content": "
Good night!
", "rawMarkdown": "\n\nGood night!", "votes": 1 }, { "id": 2162009, "postDate": "2023-02-27T23:11:03.587Z", "content": "What are you doing to those people on the bus?
", "rawMarkdown": "What are you doing to those people on the bus?", "votes": 2 } ] }, { "id": 2161272, "postDate": "2023-02-27T11:10:54.580Z", "content": "No one has power to talk 😂😂😂
", "rawMarkdown": "No one has power to talk 😂😂😂", "votes": 14, "replies": [ { "id": 2161289, "postDate": "2023-02-27T11:26:07.487Z", "content": "Haha , maybe that’s true
", "rawMarkdown": "Haha , maybe that’s true", "votes": 2, "replies": [ { "id": 2161355, "postDate": "2023-02-27T12:32:36.177Z", "content": "My summary:
\nbut because of difficulty:
\nOne small piece of advice: fasten your seat belt! 🤕😂 Good luck everyone!
", "rawMarkdown": "One small piece of advice: fasten your seat belt! 🤕😂 Good luck everyone! ", "votes": 9 }, { "id": 2161235, "postDate": "2023-02-27T10:34:13.113Z", "content": "I often see dreams about the private leaderboard rankings during the final stages of competitions, and my dreams have been accurate on few occasions. yesterday's dream was a nightmare, so I am not getting my hopes up this time :))
", "rawMarkdown": "I often see dreams about the private leaderboard rankings during the final stages of competitions, and my dreams have been accurate on few occasions. yesterday's dream was a nightmare, so I am not getting my hopes up this time :))", "votes": 8, "replies": [ { "id": 2161340, "postDate": "2023-02-27T12:15:32.007Z", "content": "I also have a dreams this week. I dreamed about someone published a very high score notebook and I dropped like hundreds… the nightmare just scared me up 😂
", "rawMarkdown": "I also have a dreams this week. I dreamed about someone published a very high score notebook and I dropped like hundreds... the nightmare just scared me up 😂", "votes": 5, "replies": [ { "id": 2161599, "postDate": "2023-02-27T15:51:30.070Z", "content": "Better start praying now:))
", "rawMarkdown": "Better start praying now:))" } ] } ] }, { "id": 2161435, "postDate": "2023-02-27T13:41:52.470Z", "content": "For me, no matter what happens I learned a lot of new things in this competition :)
", "rawMarkdown": "For me, no matter what happens I learned a lot of new things in this competition :)", "votes": 5, "replies": [ { "id": 2161465, "postDate": "2023-02-27T14:18:37.297Z", "content": "that's the spirit : )
", "rawMarkdown": "that's the spirit : )", "votes": 2 } ] }, { "id": 2161649, "postDate": "2023-02-27T16:24:58.303Z", "content": "we got radiologist to label the images. and i think the magic number is 0.65 or +/- 0.04 shakeup
", "rawMarkdown": "we got radiologist to label the images. and i think the magic number is 0.65 or +/- 0.04 shakeup", "votes": 6, "replies": [ { "id": 2161658, "postDate": "2023-02-27T16:29:58.990Z", "content": "wow…great job getting Radiologists on the team. when you say +/- 0.04, you have this for Gold zone or for all medal zones? Also, by magic number, i suppose you mean for Gold zone, right?
", "rawMarkdown": "wow...great job getting Radiologists on the team. when you say +/- 0.04, you have this for Gold zone or for all medal zones? Also, by magic number, i suppose you mean for Gold zone, right?", "votes": 1 } ] }, { "id": 2161332, "postDate": "2023-02-27T12:08:49.230Z", "content": "I think we will see high shakeup in bronze & silver zone (as they all have same LB score not much diffrence)
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3847086%2Fe52e771c50a95811e37cc6cfa51c9a43%2FScreenshot%20(442).png?generation=1677499649357374&alt=media\")
nice figure, 借你吉言(borrow your good words), i also really, really hope that there are no shakeup in the gold😆
", "rawMarkdown": "nice figure, 借你吉言(borrow your good words), i also really, really hope that there are no shakeup in the gold😆", "votes": 3, "replies": [ { "id": 2162155, "postDate": "2023-02-28T03:12:15.113Z", "content": "Congratulations to wining silver medal, as predicted there are big shakeup in silver and bronze zone.
", "rawMarkdown": "Congratulations to wining silver medal, as predicted there are big shakeup in silver and bronze zone." } ] } ] }, { "id": 2161227, "postDate": "2023-02-27T10:22:35.470Z", "content": "usually on the last day one would see Shakeup related posts from folks who know how to spot one in advance backed by good analysis. This competition haven't seen any such posts. For my team also, I am quite unsure…some of the subs look very sensitive to probability threshold and how different models are being ensembled. I am also prepared from some bit of shake…how much we will know in some hours 😅
\nAll the best to everyone!!
", "rawMarkdown": "usually on the last day one would see Shakeup related posts from folks who know how to spot one in advance backed by good analysis. This competition haven't seen any such posts. For my team also, I am quite unsure...some of the subs look very sensitive to probability threshold and how different models are being ensembled. I am also prepared from some bit of shake...how much we will know in some hours 😅\n\nAll the best to everyone!!", "votes": 6 }, { "id": 2161250, "postDate": "2023-02-27T10:49:46.033Z", "content": "I made a 0.52 sub 3 month ago, and didn't manage to improve my public LB based on improvements on CV
\nWell, we'll see :)
Same here. My best model with best threshold (based on OOF) scores 0.52 on LB. I didn't even try to emsemble multiple models. I guess we both have a chance to get a silver medal.
", "rawMarkdown": "Same here. My best model with best threshold (based on OOF) scores 0.52 on LB. I didn't even try to emsemble multiple models. I guess we both have a chance to get a silver medal." }, { "id": 2161596, "postDate": "2023-02-27T15:49:00.493Z", "content": "Same as you, and I got 0.52 just in this morning😂 I've done my best and I do got a pretty high score on my local validation, but I still didn't encounter the situation that the LB score is higher that local validation score😐😐 Anyway good luck to all you guys:)
", "rawMarkdown": "Same as you, and I got 0.52 just in this morning😂 I've done my best and I do got a pretty high score on my local validation, but I still didn't encounter the situation that the LB score is higher that local validation score😐😐 Anyway good luck to all you guys:)" } ] }, { "id": 2161485, "postDate": "2023-02-27T14:41:40.213Z", "content": "I only did 10 submissions, because i was busy( Despite that low number of submissions i got tired, anyways it was my first Kaggle competition and i actually learnt a lot from all of you. I am pretty confident that this exprience will aid me in my master`s thesis, since my theme is about breast cancer detection
", "rawMarkdown": "I only did 10 submissions, because i was busy( Despite that low number of submissions i got tired, anyways it was my first Kaggle competition and i actually learnt a lot from all of you. I am pretty confident that this exprience will aid me in my master`s thesis, since my theme is about breast cancer detection", "votes": 3 }, { "id": 2161832, "postDate": "2023-02-27T19:19:29.573Z", "content": "![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F13477216%2F2fd5baf6c7e71af8e79c8938d9e2b2a7%2Fworks.png?generation=1677525564299386&alt=media\")
High imbalanced binary data with a small ratio for the public leaderboard evaluation………👀……👀…..
\nI hope though that the best 1st place score now stays or even higher, so the community can deliver a SOTA model to the host and the important challenge area!
", "rawMarkdown": "High imbalanced binary data with a small ratio for the public leaderboard evaluation.........👀......👀.....\n\nI hope though that the best 1st place score now stays or even higher, so the community can deliver a SOTA model to the host and the important challenge area! ", "votes": 4 }, { "id": 2161299, "postDate": "2023-02-27T11:31:29.900Z", "content": "My shake forecast is ± 15-20 in gold zone, 20-50 in silver zone and 50-200 in bronze zone.
", "rawMarkdown": "My shake forecast is ± 15-20 in gold zone, 20-50 in silver zone and 50-200 in bronze zone.", "votes": 2, "replies": [ { "id": 2161302, "postDate": "2023-02-27T11:35:34.397Z", "content": "I will be happy if ur predictions come true 🤓
", "rawMarkdown": "I will be happy if ur predictions come true 🤓", "votes": 2 }, { "id": 2161325, "postDate": "2023-02-27T12:04:07.010Z", "content": "I have nearly the same thought
", "rawMarkdown": "I have nearly the same thought", "votes": 1 }, { "id": 2161617, "postDate": "2023-02-27T16:02:04.543Z", "content": "For my current LB score I'm so closed to the bronze zone according to your prediction😂
", "rawMarkdown": "For my current LB score I'm so closed to the bronze zone according to your prediction😂" } ] }, { "id": 2162039, "postDate": "2023-02-28T00:22:10.370Z", "content": "I have give up this competition for about half a month, With a 24G machine, I could only train on 1024 size with small batchsize, maybe larger size could help. But finally I found that the private leaderboard is not shake a lot for me.
", "rawMarkdown": "I have give up this competition for about half a month, With a 24G machine, I could only train on 1024 size with small batchsize, maybe larger size could help. But finally I found that the private leaderboard is not shake a lot for me.", "votes": 1, "replies": [ { "id": 2162147, "postDate": "2023-02-28T03:02:05.673Z", "content": "Congrats !! It was just opposite for me. My first submission was in final 25 days using kaggle's TPU and I faced big shakeup(+634).
", "rawMarkdown": "Congrats !! It was just opposite for me. My first submission was in final 25 days using kaggle's TPU and I faced big shakeup(+634). " } ] }, { "id": 2165294, "postDate": "2023-03-02T05:05:54.777Z", "content": "\"Why did the data scientist break up with Kaggle?
\nBecause they were overfitting the relationship and couldn't generalize to real life.\" :)
", "rawMarkdown": "\"Why did the data scientist break up with Kaggle?\n\nBecause they were overfitting the relationship and couldn't generalize to real life.\" :)", "votes": 2 }, { "id": 2162355, "postDate": "2023-02-28T06:37:15.510Z", "content": "Perhaps one more 😅
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F13477216%2Fc5c7fb09cac37bcc14d40d366cc6503b%2Fkermit.png?generation=1677566231293816&alt=media\")
Especially if the shake-up affected one very negatively.
", "rawMarkdown": "Especially if the shake-up affected one very negatively.", "votes": 1 }, { "id": 2164034, "postDate": "2023-03-01T08:59:01.137Z", "content": "Haha! That's me..
", "rawMarkdown": "Haha! That's me.." } ] }, { "id": 2161648, "postDate": "2023-02-27T16:23:22.823Z", "content": "Meme, i found co-related with this competition.
\n![\"Meme](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7557588%2Fb15dbe3610886a6a1728256093e154f3%2Finbox_3407946_b674080680317a509436948ed942db23_kaggle.jpg?generation=1677514940168404&alt=media\")
At least for me the journey was more enthralling and exiting than the result I may get, albeit a tiring one . I salute all the participants who made this journey possible for me with their suggestions and inputs. Best wishes to all.✌️
", "rawMarkdown": "At least for me the journey was more enthralling and exiting than the result I may get, albeit a tiring one . I salute all the participants who made this journey possible for me with their suggestions and inputs. Best wishes to all.✌️", "votes": 2, "replies": [ { "id": 2161638, "postDate": "2023-02-27T16:13:41.457Z", "content": "that's the spirit, the goal will always be learning.
", "rawMarkdown": "that's the spirit, the goal will always be learning.", "votes": 1 } ] }, { "id": 2161399, "postDate": "2023-02-27T13:12:59.963Z", "content": "Trust your CV
", "rawMarkdown": "Trust your CV", "votes": 2 }, { "id": 2161305, "postDate": "2023-02-27T11:37:42.860Z", "content": "I dreamed that I got 0.57. let see if the dream can forecast.
", "rawMarkdown": "I dreamed that I got 0.57. let see if the dream can forecast. ", "votes": 2, "replies": [ { "id": 2162160, "postDate": "2023-02-28T03:15:52.290Z", "content": "Opps. :) 0.57 is not even touched by anyone?
\nI still have question about host. Why same kernel scored many times, while other submissions timeout or OOM?
", "rawMarkdown": "Opps. :) 0.57 is not even touched by anyone?\n\nI still have question about host. Why same kernel scored many times, while other submissions timeout or OOM?" } ] }, { "id": 2163874, "postDate": "2023-03-01T06:10:57.133Z", "content": "Breast cancer is a major health concern for women worldwide, and early detection is critical for improving treatment outcomes. The RSNA Screening Mammography Breast Cancer Detection project is an exciting and innovative use of machine learning to improve breast cancer detection. By leveraging machine learning algorithms to analyze mammograms, the project aims to improve the accuracy and speed of breast cancer diagnosis, ultimately leading to better patient outcomes.
\nThe use of machine learning in breast cancer detection has the potential to revolutionize the field of medical imaging. With the ability to learn from large datasets and make predictions based on complex patterns, machine learning algorithms can assist radiologists in detecting even subtle signs of breast cancer that might go unnoticed by the human eye. This technology has the potential to significantly reduce the number of false negatives and false positives in mammography, leading to more accurate diagnoses and more effective treatment for patients. Overall, the RSNA Screening Mammography Breast Cancer Detection project is an exciting development that has the potential to make a significant impact on women's health.
", "rawMarkdown": "Breast cancer is a major health concern for women worldwide, and early detection is critical for improving treatment outcomes. The RSNA Screening Mammography Breast Cancer Detection project is an exciting and innovative use of machine learning to improve breast cancer detection. By leveraging machine learning algorithms to analyze mammograms, the project aims to improve the accuracy and speed of breast cancer diagnosis, ultimately leading to better patient outcomes.\n\nThe use of machine learning in breast cancer detection has the potential to revolutionize the field of medical imaging. With the ability to learn from large datasets and make predictions based on complex patterns, machine learning algorithms can assist radiologists in detecting even subtle signs of breast cancer that might go unnoticed by the human eye. This technology has the potential to significantly reduce the number of false negatives and false positives in mammography, leading to more accurate diagnoses and more effective treatment for patients. Overall, the RSNA Screening Mammography Breast Cancer Detection project is an exciting development that has the potential to make a significant impact on women's health." }, { "id": 2162111, "postDate": "2023-02-28T02:13:07.447Z", "content": "The rank first seems like counteroffensive successfully! Dramatic result!
", "rawMarkdown": "The rank first seems like counteroffensive successfully! Dramatic result!" }, { "id": 2161619, "postDate": "2023-02-27T16:03:08.193Z", "content": "Hello, I have a question: does the submission times count? i.e. More submissions lead to lower place if the scores are the same? Good luck to everyone!
", "rawMarkdown": "Hello, I have a question: does the submission times count? i.e. More submissions lead to lower place if the scores are the same? Good luck to everyone!", "replies": [ { "id": 2161624, "postDate": "2023-02-27T16:05:36.913Z", "content": "Afaik, the one with the earlier submission is ranked better in case of same scores.
", "rawMarkdown": "Afaik, the one with the earlier submission is ranked better in case of same scores." }, { "id": 2161626, "postDate": "2023-02-27T16:06:34.500Z", "content": "submission count does not count. only the two chosen one (if you choose manually) or the two with best public scores (in case you don't choose manually).
", "rawMarkdown": "submission count does not count. only the two chosen one (if you choose manually) or the two with best public scores (in case you don't choose manually)." } ] } ], "comments": [ { "id": 2161556, "author_name": "Chenglu", "author_url": "", "post_date": "2023-02-27T15:21:06.497000", "content": "as a new father, the shakeup reminds me of a song: \"the people on the bus go up and down, up and down, up and down…\"
\nbest wishes to all the ones that put in the effort in the game.. and good night(for those in GMT+8)
", "votes": 14, "replies": [ { "id": 2161849, "author_name": "Antti Isosalo", "author_url": "", "post_date": "2023-02-27T19:31:42.380000", "content": "
Good night!
", "votes": 1, "replies": [] }, { "id": 2162009, "author_name": "Ivan Aerlic", "author_url": "", "post_date": "2023-02-27T23:11:03.587000", "content": "What are you doing to those people on the bus?
", "votes": 2, "replies": [] } ] }, { "id": 2161272, "author_name": "Remek Kinas", "author_url": "", "post_date": "2023-02-27T11:10:54.580000", "content": "No one has power to talk 😂😂😂
", "votes": 14, "replies": [ { "id": 2161289, "author_name": "Chenglu", "author_url": "", "post_date": "2023-02-27T11:26:07.487000", "content": "Haha , maybe that’s true
", "votes": 2, "replies": [ { "id": 2161355, "author_name": "Remek Kinas", "author_url": "", "post_date": "2023-02-27T12:32:36.177000", "content": "My summary:
\nbut because of difficulty:
\nOne small piece of advice: fasten your seat belt! 🤕😂 Good luck everyone!
", "votes": 9, "replies": [] }, { "id": 2161235, "author_name": "Nischay Dhankhar", "author_url": "", "post_date": "2023-02-27T10:34:13.113000", "content": "I often see dreams about the private leaderboard rankings during the final stages of competitions, and my dreams have been accurate on few occasions. yesterday's dream was a nightmare, so I am not getting my hopes up this time :))
", "votes": 8, "replies": [ { "id": 2161340, "author_name": "Chenglu", "author_url": "", "post_date": "2023-02-27T12:15:32.007000", "content": "I also have a dreams this week. I dreamed about someone published a very high score notebook and I dropped like hundreds… the nightmare just scared me up 😂
", "votes": 5, "replies": [ { "id": 2161599, "author_name": "Lau2664", "author_url": "", "post_date": "2023-02-27T15:51:30.070000", "content": "Better start praying now:))
", "votes": 0, "replies": [] } ] } ] }, { "id": 2161435, "author_name": "Rasoul Mojtahedzadeh", "author_url": "", "post_date": "2023-02-27T13:41:52.470000", "content": "For me, no matter what happens I learned a lot of new things in this competition :)
", "votes": 5, "replies": [ { "id": 2161465, "author_name": "Chenglu", "author_url": "", "post_date": "2023-02-27T14:18:37.297000", "content": "that's the spirit : )
", "votes": 2, "replies": [] } ] }, { "id": 2161649, "author_name": "hengck23", "author_url": "", "post_date": "2023-02-27T16:24:58.303000", "content": "we got radiologist to label the images. and i think the magic number is 0.65 or +/- 0.04 shakeup
", "votes": 6, "replies": [ { "id": 2161658, "author_name": "sandy1112", "author_url": "", "post_date": "2023-02-27T16:29:58.990000", "content": "wow…great job getting Radiologists on the team. when you say +/- 0.04, you have this for Gold zone or for all medal zones? Also, by magic number, i suppose you mean for Gold zone, right?
", "votes": 1, "replies": [] } ] }, { "id": 2161332, "author_name": "Vipin Kumar", "author_url": "", "post_date": "2023-02-27T12:08:49.230000", "content": "I think we will see high shakeup in bronze & silver zone (as they all have same LB score not much diffrence)
\n
nice figure, 借你吉言(borrow your good words), i also really, really hope that there are no shakeup in the gold😆
", "votes": 3, "replies": [ { "id": 2162155, "author_name": "Vipin Kumar", "author_url": "", "post_date": "2023-02-28T03:12:15.113000", "content": "Congratulations to wining silver medal, as predicted there are big shakeup in silver and bronze zone.
", "votes": 0, "replies": [] } ] } ] }, { "id": 2161227, "author_name": "sandy1112", "author_url": "", "post_date": "2023-02-27T10:22:35.470000", "content": "usually on the last day one would see Shakeup related posts from folks who know how to spot one in advance backed by good analysis. This competition haven't seen any such posts. For my team also, I am quite unsure…some of the subs look very sensitive to probability threshold and how different models are being ensembled. I am also prepared from some bit of shake…how much we will know in some hours 😅
\nAll the best to everyone!!
", "votes": 6, "replies": [] }, { "id": 2161250, "author_name": "slime", "author_url": "", "post_date": "2023-02-27T10:49:46.033000", "content": "I made a 0.52 sub 3 month ago, and didn't manage to improve my public LB based on improvements on CV
\nWell, we'll see :)
Same here. My best model with best threshold (based on OOF) scores 0.52 on LB. I didn't even try to emsemble multiple models. I guess we both have a chance to get a silver medal.
", "votes": 0, "replies": [] }, { "id": 2161596, "author_name": "Lau2664", "author_url": "", "post_date": "2023-02-27T15:49:00.493000", "content": "Same as you, and I got 0.52 just in this morning😂 I've done my best and I do got a pretty high score on my local validation, but I still didn't encounter the situation that the LB score is higher that local validation score😐😐 Anyway good luck to all you guys:)
", "votes": 0, "replies": [] } ] }, { "id": 2161485, "author_name": "Gani Esenov", "author_url": "", "post_date": "2023-02-27T14:41:40.213000", "content": "I only did 10 submissions, because i was busy( Despite that low number of submissions i got tired, anyways it was my first Kaggle competition and i actually learnt a lot from all of you. I am pretty confident that this exprience will aid me in my master`s thesis, since my theme is about breast cancer detection
", "votes": 3, "replies": [] }, { "id": 2161832, "author_name": "Antti Isosalo", "author_url": "", "post_date": "2023-02-27T19:19:29.573000", "content": "
High imbalanced binary data with a small ratio for the public leaderboard evaluation………👀……👀…..
\nI hope though that the best 1st place score now stays or even higher, so the community can deliver a SOTA model to the host and the important challenge area!
", "votes": 4, "replies": [] }, { "id": 2161299, "author_name": "Gunes Evitan", "author_url": "", "post_date": "2023-02-27T11:31:29.900000", "content": "My shake forecast is ± 15-20 in gold zone, 20-50 in silver zone and 50-200 in bronze zone.
", "votes": 2, "replies": [ { "id": 2161302, "author_name": "sandy1112", "author_url": "", "post_date": "2023-02-27T11:35:34.397000", "content": "I will be happy if ur predictions come true 🤓
", "votes": 2, "replies": [] }, { "id": 2161325, "author_name": "Chenglu", "author_url": "", "post_date": "2023-02-27T12:04:07.010000", "content": "I have nearly the same thought
", "votes": 1, "replies": [] }, { "id": 2161617, "author_name": "Lau2664", "author_url": "", "post_date": "2023-02-27T16:02:04.543000", "content": "For my current LB score I'm so closed to the bronze zone according to your prediction😂
", "votes": 0, "replies": [] } ] }, { "id": 2162039, "author_name": "Mr.Fire", "author_url": "", "post_date": "2023-02-28T00:22:10.370000", "content": "I have give up this competition for about half a month, With a 24G machine, I could only train on 1024 size with small batchsize, maybe larger size could help. But finally I found that the private leaderboard is not shake a lot for me.
", "votes": 1, "replies": [ { "id": 2162147, "author_name": "Himanshu Nayal", "author_url": "", "post_date": "2023-02-28T03:02:05.673000", "content": "Congrats !! It was just opposite for me. My first submission was in final 25 days using kaggle's TPU and I faced big shakeup(+634).
", "votes": 0, "replies": [] } ] }, { "id": 2165294, "author_name": "Zubair Ahmad", "author_url": "", "post_date": "2023-03-02T05:05:54.777000", "content": "\"Why did the data scientist break up with Kaggle?
\nBecause they were overfitting the relationship and couldn't generalize to real life.\" :)
", "votes": 2, "replies": [] }, { "id": 2162355, "author_name": "Antti Isosalo", "author_url": "", "post_date": "2023-02-28T06:37:15.510000", "content": "Perhaps one more 😅
\n
Especially if the shake-up affected one very negatively.
", "votes": 1, "replies": [] }, { "id": 2164034, "author_name": "Chenglu", "author_url": "", "post_date": "2023-03-01T08:59:01.137000", "content": "Haha! That's me..
", "votes": 0, "replies": [] } ] }, { "id": 2161648, "author_name": "Dheeraj Pandey", "author_url": "", "post_date": "2023-02-27T16:23:22.823000", "content": "Meme, i found co-related with this competition.
\n
At least for me the journey was more enthralling and exiting than the result I may get, albeit a tiring one . I salute all the participants who made this journey possible for me with their suggestions and inputs. Best wishes to all.✌️
", "votes": 2, "replies": [ { "id": 2161638, "author_name": "Chenglu", "author_url": "", "post_date": "2023-02-27T16:13:41.457000", "content": "that's the spirit, the goal will always be learning.
", "votes": 1, "replies": [] } ] }, { "id": 2161399, "author_name": "Catch_Bat", "author_url": "", "post_date": "2023-02-27T13:12:59.963000", "content": "Trust your CV
", "votes": 2, "replies": [] }, { "id": 2161305, "author_name": "dragon zhang", "author_url": "", "post_date": "2023-02-27T11:37:42.860000", "content": "I dreamed that I got 0.57. let see if the dream can forecast.
", "votes": 2, "replies": [ { "id": 2162160, "author_name": "dragon zhang", "author_url": "", "post_date": "2023-02-28T03:15:52.290000", "content": "Opps. :) 0.57 is not even touched by anyone?
\nI still have question about host. Why same kernel scored many times, while other submissions timeout or OOM?
", "votes": 0, "replies": [] } ] }, { "id": 2163874, "author_name": "Ashkan Forootan", "author_url": "", "post_date": "2023-03-01T06:10:57.133000", "content": "Breast cancer is a major health concern for women worldwide, and early detection is critical for improving treatment outcomes. The RSNA Screening Mammography Breast Cancer Detection project is an exciting and innovative use of machine learning to improve breast cancer detection. By leveraging machine learning algorithms to analyze mammograms, the project aims to improve the accuracy and speed of breast cancer diagnosis, ultimately leading to better patient outcomes.
\nThe use of machine learning in breast cancer detection has the potential to revolutionize the field of medical imaging. With the ability to learn from large datasets and make predictions based on complex patterns, machine learning algorithms can assist radiologists in detecting even subtle signs of breast cancer that might go unnoticed by the human eye. This technology has the potential to significantly reduce the number of false negatives and false positives in mammography, leading to more accurate diagnoses and more effective treatment for patients. Overall, the RSNA Screening Mammography Breast Cancer Detection project is an exciting development that has the potential to make a significant impact on women's health.
", "votes": 0, "replies": [] }, { "id": 2162111, "author_name": "Qing Cha", "author_url": "", "post_date": "2023-02-28T02:13:07.447000", "content": "The rank first seems like counteroffensive successfully! Dramatic result!
", "votes": 0, "replies": [] }, { "id": 2161619, "author_name": "Xiao-Su (Frank) Hu", "author_url": "", "post_date": "2023-02-27T16:03:08.193000", "content": "Hello, I have a question: does the submission times count? i.e. More submissions lead to lower place if the scores are the same? Good luck to everyone!
", "votes": 0, "replies": [ { "id": 2161624, "author_name": "Nischay Dhankhar", "author_url": "", "post_date": "2023-02-27T16:05:36.913000", "content": "Afaik, the one with the earlier submission is ranked better in case of same scores.
", "votes": 0, "replies": [] }, { "id": 2161626, "author_name": "sandy1112", "author_url": "", "post_date": "2023-02-27T16:06:34.500000", "content": "submission count does not count. only the two chosen one (if you choose manually) or the two with best public scores (in case you don't choose manually).
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "2161197": "For me, I'm ready for some shake-down cause the submissions of this week are all about overfitting the LB.", "2161556": "as a new father, the shakeup reminds me of a song: \"the people on the bus go up and down, up and down, up and down...\"\n\nbest wishes to all the ones that put in the effort in the game.. and good night(for those in GMT+8)", "2161272": "No one has power to talk 😂😂😂", "2161517": "One small piece of advice: fasten your seat belt! 🤕😂 Good luck everyone! ", "2161235": "I often see dreams about the private leaderboard rankings during the final stages of competitions, and my dreams have been accurate on few occasions. yesterday's dream was a nightmare, so I am not getting my hopes up this time :))", "2161435": "For me, no matter what happens I learned a lot of new things in this competition :)", "2161649": "we got radiologist to label the images. and i think the magic number is 0.65 or +/- 0.04 shakeup", "2161332": "I think we will see high shakeup in **bronze & silver zone** (as they all have same LB score not much diffrence)\n\n.png?generation=1677499649357374&alt=media)", "2161227": "usually on the last day one would see Shakeup related posts from folks who know how to spot one in advance backed by good analysis. This competition haven't seen any such posts. For my team also, I am quite unsure...some of the subs look very sensitive to probability threshold and how different models are being ensembled. I am also prepared from some bit of shake...how much we will know in some hours 😅\n\nAll the best to everyone!!", "2161250": "I made a 0.52 sub 3 month ago, and didn't manage to improve my public LB based on improvements on CV\nWell, we'll see :)", "2161485": "I only did 10 submissions, because i was busy( Despite that low number of submissions i got tired, anyways it was my first Kaggle competition and i actually learnt a lot from all of you. I am pretty confident that this exprience will aid me in my master`s thesis, since my theme is about breast cancer detection", "2161832": "", "2161672": "High imbalanced binary data with a small ratio for the public leaderboard evaluation.........👀......👀.....\n\nI hope though that the best 1st place score now stays or even higher, so the community can deliver a SOTA model to the host and the important challenge area! ", "2161299": "My shake forecast is ± 15-20 in gold zone, 20-50 in silver zone and 50-200 in bronze zone.", "2162039": "I have give up this competition for about half a month, With a 24G machine, I could only train on 1024 size with small batchsize, maybe larger size could help. But finally I found that the private leaderboard is not shake a lot for me.", "2165294": "\"Why did the data scientist break up with Kaggle?\n\nBecause they were overfitting the relationship and couldn't generalize to real life.\" :)", "2162355": "Perhaps one more 😅\n", "2161648": "Meme, i found co-related with this competition. \n\n", "2161625": "At least for me the journey was more enthralling and exiting than the result I may get, albeit a tiring one . I salute all the participants who made this journey possible for me with their suggestions and inputs. Best wishes to all.✌️", "2161399": "Trust your CV", "2161305": "I dreamed that I got 0.57. let see if the dream can forecast. ", "2163874": "Breast cancer is a major health concern for women worldwide, and early detection is critical for improving treatment outcomes. The RSNA Screening Mammography Breast Cancer Detection project is an exciting and innovative use of machine learning to improve breast cancer detection. By leveraging machine learning algorithms to analyze mammograms, the project aims to improve the accuracy and speed of breast cancer diagnosis, ultimately leading to better patient outcomes.\n\nThe use of machine learning in breast cancer detection has the potential to revolutionize the field of medical imaging. With the ability to learn from large datasets and make predictions based on complex patterns, machine learning algorithms can assist radiologists in detecting even subtle signs of breast cancer that might go unnoticed by the human eye. This technology has the potential to significantly reduce the number of false negatives and false positives in mammography, leading to more accurate diagnoses and more effective treatment for patients. Overall, the RSNA Screening Mammography Breast Cancer Detection project is an exciting development that has the potential to make a significant impact on women's health.", "2162111": "The rank first seems like counteroffensive successfully! Dramatic result!", "2161619": "Hello, I have a question: does the submission times count? i.e. More submissions lead to lower place if the scores are the same? Good luck to everyone!" } }