I wonder what the people at CHAMPS predicted we would get for accuracy? With the top scores being around -3 right now the best algorithms seem to be predicting with about a ±.001 accuracy which seems good to me. Is that actually accurate enough for the people that would use these algorithms' predictions or do they need to get much more accurate?
", "messages": [ { "id": 594959, "postDate": "2019-08-08T17:23:12.113Z", "content": "I wonder what the people at CHAMPS predicted we would get for accuracy? With the top scores being around -3 right now the best algorithms seem to be predicting with about a ±.001 accuracy which seems good to me. Is that actually accurate enough for the people that would use these algorithms' predictions or do they need to get much more accurate?
", "rawMarkdown": "I wonder what the people at CHAMPS predicted we would get for accuracy? With the top scores being around -3 right now the best algorithms seem to be predicting with about a ±.001 accuracy which seems good to me. Is that actually accurate enough for the people that would use these algorithms' predictions or do they need to get much more accurate?", "votes": 3 }, { "id": 595325, "postDate": "2019-08-09T05:27:00.387Z", "content": "Already answered here:\nhttps://www.kaggle.com/c/champs-scalar-coupling/discussion/98375#latest-569312
", "rawMarkdown": "Already answered here:\nhttps://www.kaggle.com/c/champs-scalar-coupling/discussion/98375#latest-569312", "votes": 2, "replies": [ { "id": 595603, "postDate": "2019-08-09T12:50:30.783Z", "content": "oh, thank you
", "rawMarkdown": "oh, thank you" } ] }, { "id": 595147, "postDate": "2019-08-08T22:39:30.860Z", "content": "e ** - 3 is not 0.001 ;)
", "rawMarkdown": "e ** - 3 is not 0.001 ;)", "votes": 2, "replies": [ { "id": 595604, "postDate": "2019-08-09T12:51:47.133Z", "content": "I was doing log(.001) = -3, but it's just e^-a if a = your score?
", "rawMarkdown": "I was doing log(.001) = -3, but it's just e^-a if a = your score?" } ] }, { "id": 594963, "postDate": "2019-08-08T17:26:59.833Z", "content": "Good question. I was wondering the exact same question also. I'd appreciate an answer also. Thank you
", "rawMarkdown": "Good question. I was wondering the exact same question also. I'd appreciate an answer also. Thank you ", "votes": -2 }, { "id": 594966, "postDate": "2019-08-08T17:30:08.473Z", "rawMarkdown": "", "votes": -2, "isDeleted": true, "replies": [ { "id": 594968, "postDate": "2019-08-08T17:31:11.967Z", "rawMarkdown": "", "votes": -2 }, { "id": 594974, "postDate": "2019-08-08T17:35:51.887Z", "rawMarkdown": "", "votes": -1, "isDeleted": true }, { "id": 594975, "postDate": "2019-08-08T17:37:04.290Z", "rawMarkdown": "", "votes": -1 } ] }, { "id": 594964, "postDate": "2019-08-08T17:27:52.300Z", "rawMarkdown": "", "votes": 1, "isDeleted": true } ], "comments": [ { "id": 595325, "author_name": "Andrés Miguel Torrubia Sáez", "author_url": "", "post_date": "2019-08-09T05:27:00.387000", "content": "Already answered here:\nhttps://www.kaggle.com/c/champs-scalar-coupling/discussion/98375#latest-569312
", "votes": 2, "replies": [ { "id": 595603, "author_name": "corbin", "author_url": "", "post_date": "2019-08-09T12:50:30.783000", "content": "oh, thank you
", "votes": 0, "replies": [] } ] }, { "id": 595147, "author_name": "CPMP", "author_url": "", "post_date": "2019-08-08T22:39:30.860000", "content": "e ** - 3 is not 0.001 ;)
", "votes": 2, "replies": [ { "id": 595604, "author_name": "corbin", "author_url": "", "post_date": "2019-08-09T12:51:47.133000", "content": "I was doing log(.001) = -3, but it's just e^-a if a = your score?
", "votes": 0, "replies": [] } ] }, { "id": 594963, "author_name": "Jackson Shinske", "author_url": "", "post_date": "2019-08-08T17:26:59.833000", "content": "Good question. I was wondering the exact same question also. I'd appreciate an answer also. Thank you
", "votes": -2, "replies": [] }, { "id": 594966, "author_name": "", "author_url": "", "post_date": "2019-08-08T17:30:08.473000", "content": "", "votes": -2, "replies": [ { "id": 594968, "author_name": "Jackson Shinske", "author_url": "", "post_date": "2019-08-08T17:31:11.967000", "content": "", "votes": -2, "replies": [] }, { "id": 594974, "author_name": "", "author_url": "", "post_date": "2019-08-08T17:35:51.887000", "content": "", "votes": -1, "replies": [] }, { "id": 594975, "author_name": "Jackson Shinske", "author_url": "", "post_date": "2019-08-08T17:37:04.290000", "content": "", "votes": -1, "replies": [] } ] }, { "id": 594964, "author_name": "", "author_url": "", "post_date": "2019-08-08T17:27:52.300000", "content": "", "votes": 1, "replies": [] } ], "raw_markdown_by_id": { "594959": "I wonder what the people at CHAMPS predicted we would get for accuracy? With the top scores being around -3 right now the best algorithms seem to be predicting with about a ±.001 accuracy which seems good to me. Is that actually accurate enough for the people that would use these algorithms' predictions or do they need to get much more accurate?", "595325": "Already answered here:\nhttps://www.kaggle.com/c/champs-scalar-coupling/discussion/98375#latest-569312", "595147": "e ** - 3 is not 0.001 ;)", "594963": "Good question. I was wondering the exact same question also. I'd appreciate an answer also. Thank you ", "594966": "", "594964": "" } }