After the end of SIIM-Melanoma Classification, my first competition in Kaggle and one hell of a puzzle, I look forward to competing in Birdcall.
\nGiven such a short time, it would be really appreciable if people could share their ideas which might've worked for them - both successful and unsuccessful ones (latter would be much more appreciated).
\nThe open-source community works best for learning and creating better solutions - something I have learned from the abundant discussion topics in SIIM-Melanoma Classification.
Keep adding to the list. Keep learning. :)
", "messages": [ { "id": "981798", "postDate": "08/22/2020 17:30:23", "content": "After the end of SIIM-Melanoma Classification, my first competition in Kaggle and one hell of a puzzle, I look forward to competing in Birdcall.
\nGiven such a short time, it would be really appreciable if people could share their ideas which might've worked for them - both successful and unsuccessful ones (latter would be much more appreciated).
\nThe open-source community works best for learning and creating better solutions - something I have learned from the abundant discussion topics in SIIM-Melanoma Classification.
Keep adding to the list. Keep learning. :)
", "rawMarkdown": "After the end of [SIIM-Melanoma Classification](https://www.kaggle.com/c/siim-isic-melanoma-classification/overview), my first competition in Kaggle and one hell of a puzzle, I look forward to competing in Birdcall.\n\nGiven such a short time, it would be really appreciable if people could share their ideas which might've worked for them - *both successful and unsuccessful ones* **(latter would be much more appreciated).**\nThe open-source community works best for learning and creating better solutions - something I have learned from the abundant discussion topics in [SIIM-Melanoma Classification](https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion). \n\n## A few ideas on:\n#### ResNet's performance\n#### EfficientNet's performance\n#### Augmentations methods used \n\n##### Anyone tried WaveNet?\n\nKeep adding to the list. Keep learning. :)", "votes": null }, { "id": "982581", "postDate": "08/23/2020 13:59:50", "content": "A great place to get started with audio tagging is here :
\n\nThen you can give SED a look :
\n\nYou'll get more answers about your questions in the forum, but keep in mind that what worked for others may not work for you (and vice versa)
", "rawMarkdown": "A great place to get started with audio tagging is here :\n- https://www.kaggle.com/ttahara/training-birdsong-baseline-resnest50-fast\n\nThen you can give SED a look :\n- https://www.kaggle.com/hidehisaarai1213/introduction-to-sound-event-detection\n\nYou'll get more answers about your questions in the forum, but keep in mind that what worked for others may not work for you (and vice versa)", "votes": null }, { "id": "984938", "postDate": "08/25/2020 11:33:14", "content": "Sure thing @theoviel! Thanks for the pointers!👍👍
", "rawMarkdown": "Sure thing @theoviel! Thanks for the pointers!👍👍", "votes": null } ], "comments": [ { "id": 982581, "author_name": "theoviel", "author_url": "", "post_date": "08/23/2020 13:59:50", "content": "A great place to get started with audio tagging is here :
\n\nThen you can give SED a look :
\n\nYou'll get more answers about your questions in the forum, but keep in mind that what worked for others may not work for you (and vice versa)
", "votes": null, "replies": [ { "id": 984938, "author_name": "navinmundhra", "author_url": "", "post_date": "08/25/2020 11:33:14", "content": "Sure thing @theoviel! Thanks for the pointers!👍👍
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "981798": "After the end of [SIIM-Melanoma Classification](https://www.kaggle.com/c/siim-isic-melanoma-classification/overview), my first competition in Kaggle and one hell of a puzzle, I look forward to competing in Birdcall.\n\nGiven such a short time, it would be really appreciable if people could share their ideas which might've worked for them - *both successful and unsuccessful ones* **(latter would be much more appreciated).**\nThe open-source community works best for learning and creating better solutions - something I have learned from the abundant discussion topics in [SIIM-Melanoma Classification](https://www.kaggle.com/c/siim-isic-melanoma-classification/discussion). \n\n## A few ideas on:\n#### ResNet's performance\n#### EfficientNet's performance\n#### Augmentations methods used \n\n##### Anyone tried WaveNet?\n\nKeep adding to the list. Keep learning. :)", "982581": "A great place to get started with audio tagging is here :\n- https://www.kaggle.com/ttahara/training-birdsong-baseline-resnest50-fast\n\nThen you can give SED a look :\n- https://www.kaggle.com/hidehisaarai1213/introduction-to-sound-event-detection\n\nYou'll get more answers about your questions in the forum, but keep in mind that what worked for others may not work for you (and vice versa)", "984938": "Sure thing @theoviel! Thanks for the pointers!👍👍" }, "source": "meta" }