Thank you for making this contest. train_1.csv has metadata columns, such as sex, age, view category, view position, but those are missing from sample_submission_1.csv , is it possible to include that data in the sample_submission_1.csv? (I am assuming that is the correct file to use for submissions) If not, I am curious as to why they were included in the train_1.csv?
\nI do understand that sometimes, the metadata is nicely filled in, and in some institutions, they are not :(
", "messages": [ { "id": 3281674, "postDate": "2025-09-04T16:46:56.323Z", "content": "Why is metadata not provided for the test data?
", "rawMarkdown": "Why is metadata not provided for the test data?", "replies": [ { "id": 3281922, "postDate": "2025-09-05T08:35:27.170Z", "content": "Good question! I explained this just above, the metadata is only provided in train_1.csv for optional use during development, but it isn’t consistently available in test data, so it’s excluded from the submission/evaluation format.
\nFor a starting point on building models with this dataset, you can check out the CNN Baseline Model Notebook.
", "rawMarkdown": "Good question! I explained this just above, the metadata is only provided in train_1.csv for optional use during development, but it isn’t consistently available in test data, so it’s excluded from the submission/evaluation format. \n\nFor a starting point on building models with this dataset, you can check out the [CNN Baseline Model Notebook](https://www.kaggle.com/code/guntasdhanjal/simple-cnn-baseline-for-division-a).\n", "votes": 1 } ] }, { "id": 3278721, "postDate": "2025-08-30T12:39:25.183Z", "content": "Great question, and thanks for looking closely at the data!
\nThe sample_submission_1.csv only includes the target columns that we’ll actually evaluate for the competition. That’s why metadata like sex, age, view category, and view position are not part of it.
\nThose fields are included in train_1.csv because they can provide useful context during model development (if participants choose to leverage them). However, since they aren’t consistently available across all sources and because some institutions don’t record them reliably we don’t include them in the submission format. This keeps the evaluation standardized and fair for all teams.
\nSo in short:
\n✅ Metadata is there in train_1.csv for optional use.
\n❌ It’s not in sample_submission_1.csv because it won’t be evaluated.
Hope that clears it up! Thanks again for pointing it out it’s a sharp observation.
", "rawMarkdown": "Great question, and thanks for looking closely at the data! \n\nThe sample_submission_1.csv only includes the target columns that we’ll actually evaluate for the competition. That’s why metadata like sex, age, view category, and view position are not part of it.\n\nThose fields are included in train_1.csv because they can provide useful context during model development (if participants choose to leverage them). However, since they aren’t consistently available across all sources and because some institutions don’t record them reliably we don’t include them in the submission format. This keeps the evaluation standardized and fair for all teams.\n\nSo in short:\n\n✅ Metadata is there in train_1.csv for optional use.\n❌ It’s not in sample_submission_1.csv because it won’t be evaluated.\n\nHope that clears it up! Thanks again for pointing it out it’s a sharp observation." }, { "id": 3278426, "postDate": "2025-08-29T17:29:43.467Z", "content": "Thank you for making this contest. train_1.csv has metadata columns, such as sex, age, view category, view position, but those are missing from sample_submission_1.csv , is it possible to include that data in the sample_submission_1.csv? (I am assuming that is the correct file to use for submissions) If not, I am curious as to why they were included in the train_1.csv?
\nI do understand that sometimes, the metadata is nicely filled in, and in some institutions, they are not :(
", "rawMarkdown": "Thank you for making this contest. train_1.csv has metadata columns, such as sex, age, view category, view position, but those are missing from sample_submission_1.csv , is it possible to include that data in the sample_submission_1.csv? (I am assuming that is the correct file to use for submissions) If not, I am curious as to why they were included in the train_1.csv? \n\nI do understand that sometimes, the metadata is nicely filled in, and in some institutions, they are not :(" } ], "comments": [ { "id": 3281674, "author_name": "Sarthak", "author_url": "", "post_date": "2025-09-04T16:46:56.323000", "content": "Why is metadata not provided for the test data?
", "votes": 0, "replies": [ { "id": 3281922, "author_name": "Guntas Dhanjal", "author_url": "", "post_date": "2025-09-05T08:35:27.170000", "content": "Good question! I explained this just above, the metadata is only provided in train_1.csv for optional use during development, but it isn’t consistently available in test data, so it’s excluded from the submission/evaluation format.
\nFor a starting point on building models with this dataset, you can check out the CNN Baseline Model Notebook.
", "votes": 1, "replies": [] } ] }, { "id": 3278721, "author_name": "Guntas Dhanjal", "author_url": "", "post_date": "2025-08-30T12:39:25.183000", "content": "Great question, and thanks for looking closely at the data!
\nThe sample_submission_1.csv only includes the target columns that we’ll actually evaluate for the competition. That’s why metadata like sex, age, view category, and view position are not part of it.
\nThose fields are included in train_1.csv because they can provide useful context during model development (if participants choose to leverage them). However, since they aren’t consistently available across all sources and because some institutions don’t record them reliably we don’t include them in the submission format. This keeps the evaluation standardized and fair for all teams.
\nSo in short:
\n✅ Metadata is there in train_1.csv for optional use.
\n❌ It’s not in sample_submission_1.csv because it won’t be evaluated.
Hope that clears it up! Thanks again for pointing it out it’s a sharp observation.
", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "3281674": "Why is metadata not provided for the test data?", "3278721": "Great question, and thanks for looking closely at the data! \n\nThe sample_submission_1.csv only includes the target columns that we’ll actually evaluate for the competition. That’s why metadata like sex, age, view category, and view position are not part of it.\n\nThose fields are included in train_1.csv because they can provide useful context during model development (if participants choose to leverage them). However, since they aren’t consistently available across all sources and because some institutions don’t record them reliably we don’t include them in the submission format. This keeps the evaluation standardized and fair for all teams.\n\nSo in short:\n\n✅ Metadata is there in train_1.csv for optional use.\n❌ It’s not in sample_submission_1.csv because it won’t be evaluated.\n\nHope that clears it up! Thanks again for pointing it out it’s a sharp observation.", "3278426": "Thank you for making this contest. train_1.csv has metadata columns, such as sex, age, view category, view position, but those are missing from sample_submission_1.csv , is it possible to include that data in the sample_submission_1.csv? (I am assuming that is the correct file to use for submissions) If not, I am curious as to why they were included in the train_1.csv? \n\nI do understand that sometimes, the metadata is nicely filled in, and in some institutions, they are not :(" } }