There is a post about LB probing and shakeup saying that there might not be shakeup; public LB will represent final LB score.\nI don't why but I feel like there will be a huge shakeup just like PetFinder competition as both competitions are quite(maybe little bit) similar in nature. I estimate shakeup would cause(atleast) ±5 rank in current LB standing. What's your estimate??
", "messages": [ { "id": "613546", "postDate": "08/30/2019 15:49:27", "content": "There is a post about LB probing and shakeup saying that there might not be shakeup; public LB will represent final LB score.\nI don't why but I feel like there will be a huge shakeup just like PetFinder competition as both competitions are quite(maybe little bit) similar in nature. I estimate shakeup would cause(atleast) ±5 rank in current LB standing. What's your estimate??
", "rawMarkdown": "There is a [post](https://www.kaggle.com/c/aptos2019-blindness-detection/discussion/105763#latest-609621) about LB probing and shakeup saying that there might not be shakeup; public LB will represent final LB score.\nI don't why but I feel like there will be a huge shakeup just like [PetFinder](https://www.kaggle.com/c/petfinder-adoption-prediction) competition as both competitions are quite(maybe little bit) similar in nature. I estimate shakeup would cause(atleast) ±5 rank in current LB standing. What's your estimate??", "votes": null }, { "id": "613754", "postDate": "08/30/2019 19:32:50", "content": "I wouldn't be surprised if there's a little shakeup, like ±5 rank... I think it seems some people might be overfitting to the leaderboard, so they will also be greatly affected. People with robust models, and ensembles, will probably be fine. But we will see!
", "rawMarkdown": "I wouldn't be surprised if there's a little shakeup, like ±5 rank... I think it seems some people might be overfitting to the leaderboard, so they will also be greatly affected. People with robust models, and ensembles, will probably be fine. But we will see!", "votes": null } ], "comments": [ { "id": 613754, "author_name": "tanlikesmath", "author_url": "", "post_date": "08/30/2019 19:32:50", "content": "I wouldn't be surprised if there's a little shakeup, like ±5 rank... I think it seems some people might be overfitting to the leaderboard, so they will also be greatly affected. People with robust models, and ensembles, will probably be fine. But we will see!
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "613546": "There is a [post](https://www.kaggle.com/c/aptos2019-blindness-detection/discussion/105763#latest-609621) about LB probing and shakeup saying that there might not be shakeup; public LB will represent final LB score.\nI don't why but I feel like there will be a huge shakeup just like [PetFinder](https://www.kaggle.com/c/petfinder-adoption-prediction) competition as both competitions are quite(maybe little bit) similar in nature. I estimate shakeup would cause(atleast) ±5 rank in current LB standing. What's your estimate??", "613754": "I wouldn't be surprised if there's a little shakeup, like ±5 rank... I think it seems some people might be overfitting to the leaderboard, so they will also be greatly affected. People with robust models, and ensembles, will probably be fine. But we will see!" }, "source": "meta" }