{ "id": 273466, "title": "🔥 What is MGMT-value exactly? Can you help me?🔥 ", "url": "/competitions/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/273466", "author_name": "", "post_date": "2021-09-21T07:23:59.244065600Z", "votes": 1, "comment_count": 6, "views": 0, "content": "

Hi everyone. I didn't understand MGMT-value what is exactly. we trained models with one neuron in last layer and sigmoid activation function. IN train.csv file each patient have label 0 or 1 but when we predict test data get probability between 0 and 1 . what is exactly mean?
\nwhen we submit prediction how our model will be evaluated?
\nThanks foe your help.🔥

", "messages": [ { "id": "1518906", "postDate": "09/21/2021 07:23:59", "content": "

Hi everyone. I didn't understand MGMT-value what is exactly. we trained models with one neuron in last layer and sigmoid activation function. IN train.csv file each patient have label 0 or 1 but when we predict test data get probability between 0 and 1 . what is exactly mean?
\nwhen we submit prediction how our model will be evaluated?
\nThanks foe your help.🔥

", "rawMarkdown": "**Hi everyone**. I didn't understand MGMT-value what is exactly. we trained models with one neuron in last layer and sigmoid activation function. IN train.csv file each patient have label 0 or 1 but when we predict test data get probability between 0 and 1 . what is exactly mean? \nwhen we submit prediction how our model will be evaluated?\nThanks foe your help.🔥", "votes": null }, { "id": "1519051", "postDate": "09/21/2021 09:32:18", "content": "

You should refer to these two threads:
\n1)https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/264861
\n2)https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/253093

", "rawMarkdown": "You should refer to these two threads: \n1)https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/264861\n2)https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/253093", "votes": null }, { "id": "1520124", "postDate": "09/22/2021 06:28:40", "content": "

Thanks. I saw this link but i can't understand what is the different between 0 and 1 label for each id .
\ncan you tell me about this?

", "rawMarkdown": "Thanks. I saw this link but i can't understand what is the different between 0 and 1 label for each id .\ncan you tell me about this?", "votes": null }, { "id": "1520133", "postDate": "09/22/2021 06:41:19", "content": "
\n

train_labels.csv - file containing the target MGMT_value for each subject in the training data (e.g. the presence of MGMT promoter methylation)

\n
\n

I think that if MGMT promoter methylation is present then MGMT_value will be equal to 1 otherwise 0.

", "rawMarkdown": "> train_labels.csv - file containing the target MGMT_value for each subject in the training data (e.g. the presence of MGMT promoter methylation)\n\nI think that if MGMT promoter methylation is present then MGMT_value will be equal to 1 otherwise 0.", "votes": null }, { "id": "1520154", "postDate": "09/22/2021 06:56:41", "content": "

how we can find MGMT promoter methylation in 3D images after visualization?
\nAll images contain tumor but MGMT promoter show what exactly in images with label 1 and 0?

", "rawMarkdown": "how we can find MGMT promoter methylation in 3D images after visualization?\nAll images contain tumor but MGMT promoter show what exactly in images with label 1 and 0?", "votes": null }, { "id": "1520202", "postDate": "09/22/2021 07:36:32", "content": "

For finding MGMT promoter methylation or if it is even possible or not check out this thread:https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/271955
\nI think this thread will be useful for your second question: https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/267750

", "rawMarkdown": "For finding MGMT promoter methylation or if it is even possible or not check out this thread:https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/271955\nI think this thread will be useful for your second question: https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/267750", "votes": null }, { "id": "1520263", "postDate": "09/22/2021 08:30:03", "content": "

Thanks a lot

", "rawMarkdown": "Thanks a lot", "votes": null } ], "comments": [ { "id": 1519051, "author_name": "susnato", "author_url": "", "post_date": "09/21/2021 09:32:18", "content": "

You should refer to these two threads:
\n1)https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/264861
\n2)https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/253093

", "votes": null, "replies": [ { "id": 1520124, "author_name": "mohammadhosein1998", "author_url": "", "post_date": "09/22/2021 06:28:40", "content": "

Thanks. I saw this link but i can't understand what is the different between 0 and 1 label for each id .
\ncan you tell me about this?

", "votes": null, "replies": [] }, { "id": 1520133, "author_name": "susnato", "author_url": "", "post_date": "09/22/2021 06:41:19", "content": "
\n

train_labels.csv - file containing the target MGMT_value for each subject in the training data (e.g. the presence of MGMT promoter methylation)

\n
\n

I think that if MGMT promoter methylation is present then MGMT_value will be equal to 1 otherwise 0.

", "votes": null, "replies": [] }, { "id": 1520154, "author_name": "mohammadhosein1998", "author_url": "", "post_date": "09/22/2021 06:56:41", "content": "

how we can find MGMT promoter methylation in 3D images after visualization?
\nAll images contain tumor but MGMT promoter show what exactly in images with label 1 and 0?

", "votes": null, "replies": [] }, { "id": 1520202, "author_name": "susnato", "author_url": "", "post_date": "09/22/2021 07:36:32", "content": "

For finding MGMT promoter methylation or if it is even possible or not check out this thread:https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/271955
\nI think this thread will be useful for your second question: https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/267750

", "votes": null, "replies": [] }, { "id": 1520263, "author_name": "mohammadhosein1998", "author_url": "", "post_date": "09/22/2021 08:30:03", "content": "

Thanks a lot

", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1518906": "**Hi everyone**. I didn't understand MGMT-value what is exactly. we trained models with one neuron in last layer and sigmoid activation function. IN train.csv file each patient have label 0 or 1 but when we predict test data get probability between 0 and 1 . what is exactly mean? \nwhen we submit prediction how our model will be evaluated?\nThanks foe your help.🔥", "1519051": "You should refer to these two threads: \n1)https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/264861\n2)https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/253093", "1520124": "Thanks. I saw this link but i can't understand what is the different between 0 and 1 label for each id .\ncan you tell me about this?", "1520133": "> train_labels.csv - file containing the target MGMT_value for each subject in the training data (e.g. the presence of MGMT promoter methylation)\n\nI think that if MGMT promoter methylation is present then MGMT_value will be equal to 1 otherwise 0.", "1520154": "how we can find MGMT promoter methylation in 3D images after visualization?\nAll images contain tumor but MGMT promoter show what exactly in images with label 1 and 0?", "1520202": "For finding MGMT promoter methylation or if it is even possible or not check out this thread:https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/271955\nI think this thread will be useful for your second question: https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/267750", "1520263": "Thanks a lot" }, "source": "meta" }