Hi,\nI'm quite familiar with machine learning, but only in cases where I'm dealing with tabular data. I'm not sure how to start and work with image analysis competition (openCV?, CNN?). Can someone point me to related MOOC, book or online tutorial? My preference is for something technical and quick to implement rather than over-talked and theoretical, but I do not despise anything educational.
\n\nThanks and good luck to everyone!
", "messages": [ { "id": "152770", "postDate": "12/28/2016 13:07:01", "content": "Hi,\nI'm quite familiar with machine learning, but only in cases where I'm dealing with tabular data. I'm not sure how to start and work with image analysis competition (openCV?, CNN?). Can someone point me to related MOOC, book or online tutorial? My preference is for something technical and quick to implement rather than over-talked and theoretical, but I do not despise anything educational.
\n\nThanks and good luck to everyone!
", "rawMarkdown": "Hi,\r\nI'm quite familiar with machine learning, but only in cases where I'm dealing with tabular data. I'm not sure how to start and work with image analysis competition (openCV?, CNN?). Can someone point me to related MOOC, book or online tutorial? My preference is for something technical and quick to implement rather than over-talked and theoretical, but I do not despise anything educational.\r\n\r\nThanks and good luck to everyone!", "votes": null }, { "id": "152775", "postDate": "12/28/2016 13:23:12", "content": "Hi QuantumDamage,
\n\nThere were previous competitions of this nature, I find it helpful to go to their topics and read solutions and kernels. See Right Whale Recognition for example.
\n\nUseful Compilation made by SRK
\n\nSome others (notice the required level of experience varies a lot):
\n\nLeaf Classification\nhttps://www.kaggle.com/c/leaf-classification
\n\nDigit Recognizer\nhttps://www.kaggle.com/c/digit-recognizer
\n\nDogs vs. Cats Redux: Kernels Edition\nhttps://www.kaggle.com/c/dogs-vs-cats-redux-kernels-edition
\n\nFacial Keypoints Detection\nhttps://www.kaggle.com/c/facial-keypoints-detection
\n\nRight Whale Recognition\nhttps://www.kaggle.com/c/noaa-right-whale-recognition
\n\np.s.\nthat's a lot of reading to do :-) But hopefully we'll have a lot of time with NNs needing several hours per Epoch such as in the article shared earlier. I don't have a GPU yet soooo
", "rawMarkdown": "Hi QuantumDamage,\r\n\r\nThere were previous competitions of this nature, I find it helpful to go to their topics and read solutions and kernels. See [Right Whale Recognition][1] for example.\r\n\r\nUseful [Compilation made by SRK][2]\r\n\r\nSome others (notice the required level of experience varies a lot):\r\n\r\nLeaf Classification\r\nhttps://www.kaggle.com/c/leaf-classification\r\n\r\nDigit Recognizer\r\nhttps://www.kaggle.com/c/digit-recognizer\r\n\r\nDogs vs. Cats Redux: Kernels Edition\r\nhttps://www.kaggle.com/c/dogs-vs-cats-redux-kernels-edition\r\n\r\nFacial Keypoints Detection\r\nhttps://www.kaggle.com/c/facial-keypoints-detection\r\n\r\nRight Whale Recognition\r\nhttps://www.kaggle.com/c/noaa-right-whale-recognition\r\n\r\n\r\n [1]: https://www.kaggle.com/c/facial-keypoints-detection\r\n [2]: https://www.kaggle.com/c/the-nature-conservancy-fisheries-monitoring/forums/t/25439/interesting-solutions-from-the-past?forumMessageId=144563#post144563\r\n\r\np.s.\r\nthat's a lot of reading to do :-) But hopefully we'll have a lot of time with NNs needing several hours per Epoch such as in the article shared earlier. I don't have a GPU yet soooo", "votes": null }, { "id": "152778", "postDate": "12/28/2016 13:38:20", "content": "Ah, forgot to mention... some reading applicable to satellite imagery:\nhttps://www.cs.toronto.edu/~vmnih/docs/Mnih_Volodymyr_PhD_Thesis.pdf
", "rawMarkdown": "Ah, forgot to mention... some reading applicable to satellite imagery:\r\nhttps://www.cs.toronto.edu/~vmnih/docs/Mnih_Volodymyr_PhD_Thesis.pdf", "votes": null }, { "id": "152795", "postDate": "12/28/2016 15:02:11", "content": "I would highly recommend this MOOC\nhttp://course.fast.ai/lessons/lessons.html\nExcellent lessons and explanations of CNN. Especially lesson #3 where the author shows few tricks on how to set up and optimize a CNN.
", "rawMarkdown": "I would highly recommend this MOOC\r\nhttp://course.fast.ai/lessons/lessons.html\r\nExcellent lessons and explanations of CNN. Especially lesson #3 where the author shows few tricks on how to set up and optimize a CNN.", "votes": null } ], "comments": [ { "id": 152775, "author_name": "aamaia", "author_url": "", "post_date": "12/28/2016 13:23:12", "content": "Hi QuantumDamage,
\n\nThere were previous competitions of this nature, I find it helpful to go to their topics and read solutions and kernels. See Right Whale Recognition for example.
\n\nUseful Compilation made by SRK
\n\nSome others (notice the required level of experience varies a lot):
\n\nLeaf Classification\nhttps://www.kaggle.com/c/leaf-classification
\n\nDigit Recognizer\nhttps://www.kaggle.com/c/digit-recognizer
\n\nDogs vs. Cats Redux: Kernels Edition\nhttps://www.kaggle.com/c/dogs-vs-cats-redux-kernels-edition
\n\nFacial Keypoints Detection\nhttps://www.kaggle.com/c/facial-keypoints-detection
\n\nRight Whale Recognition\nhttps://www.kaggle.com/c/noaa-right-whale-recognition
\n\np.s.\nthat's a lot of reading to do :-) But hopefully we'll have a lot of time with NNs needing several hours per Epoch such as in the article shared earlier. I don't have a GPU yet soooo
", "votes": null, "replies": [] }, { "id": 152778, "author_name": "aamaia", "author_url": "", "post_date": "12/28/2016 13:38:20", "content": "Ah, forgot to mention... some reading applicable to satellite imagery:\nhttps://www.cs.toronto.edu/~vmnih/docs/Mnih_Volodymyr_PhD_Thesis.pdf
", "votes": null, "replies": [] }, { "id": 152795, "author_name": "sakvaua", "author_url": "", "post_date": "12/28/2016 15:02:11", "content": "I would highly recommend this MOOC\nhttp://course.fast.ai/lessons/lessons.html\nExcellent lessons and explanations of CNN. Especially lesson #3 where the author shows few tricks on how to set up and optimize a CNN.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "152770": "Hi,\r\nI'm quite familiar with machine learning, but only in cases where I'm dealing with tabular data. I'm not sure how to start and work with image analysis competition (openCV?, CNN?). Can someone point me to related MOOC, book or online tutorial? My preference is for something technical and quick to implement rather than over-talked and theoretical, but I do not despise anything educational.\r\n\r\nThanks and good luck to everyone!", "152775": "Hi QuantumDamage,\r\n\r\nThere were previous competitions of this nature, I find it helpful to go to their topics and read solutions and kernels. See [Right Whale Recognition][1] for example.\r\n\r\nUseful [Compilation made by SRK][2]\r\n\r\nSome others (notice the required level of experience varies a lot):\r\n\r\nLeaf Classification\r\nhttps://www.kaggle.com/c/leaf-classification\r\n\r\nDigit Recognizer\r\nhttps://www.kaggle.com/c/digit-recognizer\r\n\r\nDogs vs. Cats Redux: Kernels Edition\r\nhttps://www.kaggle.com/c/dogs-vs-cats-redux-kernels-edition\r\n\r\nFacial Keypoints Detection\r\nhttps://www.kaggle.com/c/facial-keypoints-detection\r\n\r\nRight Whale Recognition\r\nhttps://www.kaggle.com/c/noaa-right-whale-recognition\r\n\r\n\r\n [1]: https://www.kaggle.com/c/facial-keypoints-detection\r\n [2]: https://www.kaggle.com/c/the-nature-conservancy-fisheries-monitoring/forums/t/25439/interesting-solutions-from-the-past?forumMessageId=144563#post144563\r\n\r\np.s.\r\nthat's a lot of reading to do :-) But hopefully we'll have a lot of time with NNs needing several hours per Epoch such as in the article shared earlier. I don't have a GPU yet soooo", "152778": "Ah, forgot to mention... some reading applicable to satellite imagery:\r\nhttps://www.cs.toronto.edu/~vmnih/docs/Mnih_Volodymyr_PhD_Thesis.pdf", "152795": "I would highly recommend this MOOC\r\nhttp://course.fast.ai/lessons/lessons.html\r\nExcellent lessons and explanations of CNN. Especially lesson #3 where the author shows few tricks on how to set up and optimize a CNN." }, "source": "meta" }