Hello,
\nI would like to know that whether this is a multi-label classification or a simple multi-class classification?
Hello,
\nI would like to know that whether this is a multi-label classification or a simple multi-class classification?
This is a multi label problem, unlike last year.
", "rawMarkdown": "This is a multi label problem, unlike last year.", "votes": 4 }, { "id": 1240454, "postDate": "2021-03-16T12:24:33.090Z", "content": "What I understood is that, some images have multiple labels and among them some images are labelled as complex and some are labelled as complex as well as some disease names separated by delimiter and some have multiple labels without complex.
Reference :
\n 8002cb321f8bfcdf.jpg scab frog_eye_leaf_spot complex\n\n 801f78399a44e7af.jpg complex\n\n 80769797ce42f658.jpg scab frog_eye_leaf_spot\n@mohammadasimbluemoon Even I have the same query . From last year notebook , I went through Tarun EDA , only 5 percent of plants have multiple diseases and many have single diseases only . But this year in the description , they have mentioned they have added additional diseases .
", "rawMarkdown": "@mohammadasimbluemoon Even I have the same query . From last year notebook , I went through Tarun EDA , only 5 percent of plants have multiple diseases and many have single diseases only . But this year in the description , they have mentioned they have added additional diseases . ", "votes": 2, "replies": [ { "id": 1239986, "postDate": "2021-03-16T07:07:40.767Z", "content": "Thanks for giving me reply. Yes you are right, there is still some confusion regarding the number of diseases and the type of classification. I think we have to treat it as a multi-label problem as there are multiple diseases in a single sample. currently i have found the following labels…
\nlabel_names=['healthy', 'scab', 'frog_eye_leaf_spot', 'cider_apple_rust', 'complex', 'powdery_mildew', 'rust']
These are the 7 labels and the dataset contains either individual labels or their combination. So instead of another dedicated class for combination of diseases like in the 2020 dataset and hence would act as multiclass, but this time it is more detailed hence it is multilabel.
", "rawMarkdown": "Thanks for giving me reply. Yes you are right, there is still some confusion regarding the number of diseases and the type of classification. I think we have to treat it as a multi-label problem as there are multiple diseases in a single sample. currently i have found the following labels...\nlabel_names=['healthy', 'scab', 'frog_eye_leaf_spot', 'cider_apple_rust', 'complex', 'powdery_mildew', 'rust']\n\nThese are the 7 labels and the dataset contains either individual labels or their combination. So instead of another dedicated class for combination of diseases like in the 2020 dataset and hence would act as multiclass, but this time it is more detailed hence it is multilabel.\n", "votes": 1 } ] }, { "id": 1239982, "postDate": "2021-03-16T07:06:22.243Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 1240899, "author_name": "Sohier Dane", "author_url": "", "post_date": "2021-03-16T18:03:27.140000", "content": "This is a multi label problem, unlike last year.
", "votes": 4, "replies": [] }, { "id": 1240454, "author_name": "Shanmukh", "author_url": "", "post_date": "2021-03-16T12:24:33.090000", "content": "What I understood is that, some images have multiple labels and among them some images are labelled as complex and some are labelled as complex as well as some disease names separated by delimiter and some have multiple labels without complex.
Reference :
\n 8002cb321f8bfcdf.jpg scab frog_eye_leaf_spot complex\n\n 801f78399a44e7af.jpg complex\n\n 80769797ce42f658.jpg scab frog_eye_leaf_spot\n@mohammadasimbluemoon Even I have the same query . From last year notebook , I went through Tarun EDA , only 5 percent of plants have multiple diseases and many have single diseases only . But this year in the description , they have mentioned they have added additional diseases .
", "votes": 2, "replies": [ { "id": 1239986, "author_name": "MohammadAsim@bluemoon", "author_url": "", "post_date": "2021-03-16T07:07:40.767000", "content": "Thanks for giving me reply. Yes you are right, there is still some confusion regarding the number of diseases and the type of classification. I think we have to treat it as a multi-label problem as there are multiple diseases in a single sample. currently i have found the following labels…
\nlabel_names=['healthy', 'scab', 'frog_eye_leaf_spot', 'cider_apple_rust', 'complex', 'powdery_mildew', 'rust']
These are the 7 labels and the dataset contains either individual labels or their combination. So instead of another dedicated class for combination of diseases like in the 2020 dataset and hence would act as multiclass, but this time it is more detailed hence it is multilabel.
", "votes": 1, "replies": [] } ] }, { "id": 1239982, "author_name": "", "author_url": "", "post_date": "2021-03-16T07:06:22.243000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "1239931": "Hello,\nI would like to know that whether this is a multi-label classification or a simple multi-class classification?", "1240899": "This is a multi label problem, unlike last year.", "1240454": "What I understood is that, some images have multiple labels and among them some images are labelled as `complex` and some are labelled as `complex` as well as some disease names separated by `delimiter` and some have multiple labels without `complex`.\n\nReference :\n\n 8002cb321f8bfcdf.jpg scab frog_eye_leaf_spot complex\n\n 801f78399a44e7af.jpg complex\n\n 80769797ce42f658.jpg scab frog_eye_leaf_spot", "1239945": "@mohammadasimbluemoon Even I have the same query . From last year notebook , I went through Tarun EDA , only 5 percent of plants have multiple diseases and many have single diseases only . But this year in the description , they have mentioned they have added additional diseases . ", "1239982": "" } }