I saw the sample submission file, and i notice that it has more than one target, then i think this is a multi-label classification task, am i wrong?. Maybe this question is too easy for some experienced kagglers, however i still have my doubts.
", "messages": [ { "id": 1112725, "postDate": "2020-12-14T20:45:55.817Z", "content": "I saw the sample submission file, and i notice that it has more than one target, then i think this is a multi-label classification task, am i wrong?. Maybe this question is too easy for some experienced kagglers, however i still have my doubts.
", "rawMarkdown": "I saw the sample submission file, and i notice that it has more than one target, then i think this is a **multi-label classification** task, am i wrong?. Maybe this question is too easy for some experienced kagglers, however i still have my doubts.", "votes": 9 }, { "id": 1112784, "postDate": "2020-12-14T22:15:54.853Z", "content": "Hi!
\nThis is the competition host. Thanks for participating in the competition. To confirm, this is a multi label binary classification task.
\nRegards,
\nJarrel
\nRANZCR Challenge Organising Team
Thank you so much for answer.
", "rawMarkdown": "Thank you so much for answer." }, { "id": 1147529, "postDate": "2021-01-10T15:10:47.690Z", "content": "Hi, is it possible to have multiple annotation for the same group of catheters? For example can one scan be simultaneously annotated more than one of the groups:
\nNGT - Abnormal NGT - Borderline NGT - Incompletely Imaged NGT - Normal\nor \nCVC - Abnormal CVC - Borderline CVC - Normal\nIf no, then there are 45 scans with multiple NGT annotations and 3575 with multiple CVC annotations.
", "rawMarkdown": "Hi, is it possible to have multiple annotation for the same group of catheters? For example can one scan be simultaneously annotated more than one of the groups:\n\n```\nNGT - Abnormal\tNGT - Borderline\tNGT - Incompletely Imaged\tNGT - Normal\nor \nCVC - Abnormal\tCVC - Borderline\tCVC - Normal\n```\n\nIf no, then there are 45 scans with multiple NGT annotations and 3575 with multiple CVC annotations.\n\n\n " } ] }, { "id": 1112769, "postDate": "2020-12-14T22:01:56.303Z", "content": "The train.csv file contains classes that are labeled more than once for some of the rows. So, yes, this is a multilabel classification problem. The co-occurance of targets are available in this EDA notebook.
", "rawMarkdown": "The train.csv file contains classes that are labeled more than once for some of the rows. So, yes, this is a multilabel classification problem. The co-occurance of targets are available in [this](https://www.kaggle.com/tolgadincer/ranzcr-clip-eda) EDA notebook.", "votes": 2, "replies": [ { "id": 1112771, "postDate": "2020-12-14T22:03:11.797Z", "content": "Thanks for answer.
", "rawMarkdown": "Thanks for answer." } ] }, { "id": 1116457, "postDate": "2020-12-17T07:28:50.033Z", "content": "Yup sure. There are several binary features to predict, thus you would better go with NNs for multi-label classification
", "rawMarkdown": "Yup sure. There are several binary features to predict, thus you would better go with NNs for multi-label classification " }, { "id": 1112773, "postDate": "2020-12-14T22:03:30.543Z", "content": "Yes it is multi-label classification
", "rawMarkdown": "Yes it is multi-label classification" }, { "id": 1116463, "postDate": "2020-12-17T07:33:33.047Z", "rawMarkdown": "", "votes": -1, "isDeleted": true } ], "comments": [ { "id": 1112784, "author_name": "Jarrel Seah", "author_url": "", "post_date": "2020-12-14T22:15:54.853000", "content": "Hi!
\nThis is the competition host. Thanks for participating in the competition. To confirm, this is a multi label binary classification task.
\nRegards,
\nJarrel
\nRANZCR Challenge Organising Team
Thank you so much for answer.
", "votes": 0, "replies": [] }, { "id": 1147529, "author_name": "Kerem Turgutlu", "author_url": "", "post_date": "2021-01-10T15:10:47.690000", "content": "Hi, is it possible to have multiple annotation for the same group of catheters? For example can one scan be simultaneously annotated more than one of the groups:
\nNGT - Abnormal NGT - Borderline NGT - Incompletely Imaged NGT - Normal\nor \nCVC - Abnormal CVC - Borderline CVC - Normal\nIf no, then there are 45 scans with multiple NGT annotations and 3575 with multiple CVC annotations.
", "votes": 0, "replies": [] } ] }, { "id": 1112769, "author_name": "Tolga", "author_url": "", "post_date": "2020-12-14T22:01:56.303000", "content": "The train.csv file contains classes that are labeled more than once for some of the rows. So, yes, this is a multilabel classification problem. The co-occurance of targets are available in this EDA notebook.
", "votes": 2, "replies": [ { "id": 1112771, "author_name": "Hiram Coria 🧬", "author_url": "", "post_date": "2020-12-14T22:03:11.797000", "content": "Thanks for answer.
", "votes": 0, "replies": [] } ] }, { "id": 1116457, "author_name": "Elvin Aghammadzada", "author_url": "", "post_date": "2020-12-17T07:28:50.033000", "content": "Yup sure. There are several binary features to predict, thus you would better go with NNs for multi-label classification
", "votes": 0, "replies": [] }, { "id": 1112773, "author_name": "Saayed Alam", "author_url": "", "post_date": "2020-12-14T22:03:30.543000", "content": "Yes it is multi-label classification
", "votes": 0, "replies": [] }, { "id": 1116463, "author_name": "", "author_url": "", "post_date": "2020-12-17T07:33:33.047000", "content": "", "votes": -1, "replies": [] } ], "raw_markdown_by_id": { "1112725": "I saw the sample submission file, and i notice that it has more than one target, then i think this is a **multi-label classification** task, am i wrong?. Maybe this question is too easy for some experienced kagglers, however i still have my doubts.", "1112784": "Hi!\n\nThis is the competition host. Thanks for participating in the competition. To confirm, this is a multi label binary classification task. \n\nRegards, \nJarrel\nRANZCR Challenge Organising Team", "1112769": "The train.csv file contains classes that are labeled more than once for some of the rows. So, yes, this is a multilabel classification problem. The co-occurance of targets are available in [this](https://www.kaggle.com/tolgadincer/ranzcr-clip-eda) EDA notebook.", "1116457": "Yup sure. There are several binary features to predict, thus you would better go with NNs for multi-label classification ", "1112773": "Yes it is multi-label classification", "1116463": "" } }