{ "id": 186349, "title": "Possible resources for you and mostly what I have learned thus far.", "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/186349", "author_name": "", "post_date": "2020-09-24T07:15:35.112779200Z", "votes": 11, "comment_count": 10, "views": 0, "content": "

To my fellow Kagglers like myself,

\n

There are a lot of resources in this post which may be of interest to you. I hope you enjoy. Thank you.

\n

To the creators of this competition,

\n

I want to thank you up front for the creation of this competition. I learned a lot as I attempted to do this OSIC Pulmonary Fibrosis competition. They are as follows:

\n
    \n

  1. I got more comfortable with the Jupyter environment.

    2. \n

  2. I learned that combining kernels is not as simple as it sounds. I combined two kernels, ran it, and then had problems. I learned that code is dependent upon the code just preceding it.

    3. \n

  3. I learned about auto machine learning software. I did a free trial run of RapidMiner. This is a good software program. However, the software program that I have come to really enjoy was Orange. I like Orange because it is part of the Anaconda software package. I also learned the strength and weaknesses of Orange. Here is the youtube channel on Orange:

    4. \n
\n

https://www.youtube.com/channel/UClKKWBe2SCAEyv7ZNGhIe4g

\n
    \n
  1. I learned that the auto-machine learning software Accord.Net does its machine learning projects in C#. Because of this, I have decided that I am not going to utilize this software. Here is the link to that information though:
    2. \n
\n

https://github.com/accord-net/framework/wiki/Getting-started

\n
    \n
  1. I learned that I needed to learn tensorflow. I read half of the following book on it:
    2. \n
\n

https://www.amazon.com/Hands-Machine-Learning-Scikit-Learn-TensorFlow-ebook-dp-B07XGF2G87/dp/B07XGF2G87/ref=mt_other?_encoding=UTF8&me=&qid=

\n
    \n
  1. I learned how to install tensorflow in the anaconda package.
    2. \n
\n

https://towardsdatascience.com/https-medium-com-ekapope-v-install-tensorflow-and-keras-using-anaconda-navigator-without-command-line-b0bc41dbd038

\n
    \n
  1. To learn about Tensorflow, I watched the following Youtube video:
    2. \n
\n

https://www.youtube.com/watch?v=VwVg9jCtqaU

\n

Links associated with the above video:
\nhttps://codelabs.developers.google.com/codelabs/tensorflow-lab3-convolutions/index.html?index=..%2F..index#3

\n

https://colab.research.google.com/github/lmoroney/io19/blob/master/Zero%20to%20Hero/Rock-Paper-Scissors.ipynb

\n

https://www.tensorflow.org/tfx

\n
    \n
  1. I learned of the google developer codelabs websites. These websites are not just useful for learning Tensorflow, but they are good for learning many other languages and issues. They are the following:
    2. \n
\n

https://codelabs.developers.google.com/
\nhttps://codelabs.developers.google.com/learn-tensorflow/

\n
    \n
  1. I learned of the many models available on the tensorflow.org website. I learned that these models do not have python code embedded in them. I also learned that I needed to download these models as if they were a dataset.
    2. \n
\n

My discussion post on Kaggle when I learned of the above information:
\nhttps://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/discussion/181247

\n

The tensorflow hub where you can find models for images:
\nhttps://tfhub.dev/s?module-type=image-augmentation,image-classification,image-feature-vector,image-generator,image-object-detection,image-others,image-style-transfer,image-rnn-agent

\n
    \n
  1. I learned a little bit about TensorBoard
    2. \n
\n

https://www.tensorflow.org/tensorboard

\n
    \n
  1. I learned a little bit about the \"WhatIf\" model for tensorflow.
    2. \n
\n

https://pair-code.github.io/what-if-tool/

\n
    \n
  1. I learned that there was a certificate program (or test) to learn tensorflow. I plan on earning that certificate at some point in time.
    2. \n
\n

https://www.tensorflow.org/certificate

\n
    \n
  1. I learned of Pycharm and downloaded that program on my computer.
    2. \n
\n

https://www.jetbrains.com/pycharm/

\n
    \n
  1. I learned a little bit about Fast AI. In fact, I bought an O'Reilly book on it.
    2. \n
\n

https://www.amazon.com/Deep-Learning-Coders-fastai-PyTorch-ebook-dp-B08C2KM7NR/dp/B08C2KM7NR/ref=mt_other?_encoding=UTF8&me=&qid=1600926409

\n
    \n
  1. I learned a little bit about Hounsfield units, dicom files, and pydicom:
    2. \n
\n

https://pydicom.github.io/pydicom/dev/index.html
\nhttps://pydicom.github.io/pydicom/stable/index.html#
\nhttps://dicom.innolitics.com/ciods
\nhttps://pydicom.github.io/pydicom/0.9/working_with_pixel_data.html

\n
    \n
  1. I am learning about the kaggle API and determining if it is a good thing to use or not:
    2. \n
\n

https://github.com/Kaggle/kaggle-api

\n
    \n
  1. I also learned that the first-place team for another competition utilized many models to find the best solution.
    2. \n
\n

https://github.com/haqishen/SIIM-ISIC-Melanoma-Classification-1st-Place-Solution

\n
    \n
  1. I read Andrew Ng's book “Machine Learning Yearning.” Github found at following location:
    2. \n
\n

https://github.com/ajaymache/machine-learning-yearning

\n
    \n

  1. I learned a lot from others, but I also learned that I need to develop kernels of my own because other people’s kernel’s simply may not work, or they may have errors in them. In fact, I even recall copying a kernel and then running it and there were a lot of error codes which resulted after running them.

    2. \n

  2. I really like the following resource which gives python code examples which work with dicom files:

    3. \n
\n

https://www.programcreek.com/python/example/97517/dicom.read_file

\n

For above website I really like the last example because it converts the dicom file into an np array. My thinking was that if I convert the files into an np array that I could do

\n
    \n
  1. I also like the following resources because it teaches you how to use the models from the tensorflow hub:
    2. \n
\n

https://www.tensorflow.org/tutorials/images/transfer_learning_with_hub

\n

https://www.tensorflow.org/api_docs/python/tf/keras/utils/get_file

\n

Anyway, I hope that this post helps someone for this competition. I still have a lot to learn. Please let me know what your thoughts are if you have any.

\n

Thank you,
\nBrian

", "messages": [ { "id": "1024864", "postDate": "09/24/2020 07:15:35", "content": "

To my fellow Kagglers like myself,

\n

There are a lot of resources in this post which may be of interest to you. I hope you enjoy. Thank you.

\n

To the creators of this competition,

\n

I want to thank you up front for the creation of this competition. I learned a lot as I attempted to do this OSIC Pulmonary Fibrosis competition. They are as follows:

\n
    \n

  1. I got more comfortable with the Jupyter environment.

    2. \n

  2. I learned that combining kernels is not as simple as it sounds. I combined two kernels, ran it, and then had problems. I learned that code is dependent upon the code just preceding it.

    3. \n

  3. I learned about auto machine learning software. I did a free trial run of RapidMiner. This is a good software program. However, the software program that I have come to really enjoy was Orange. I like Orange because it is part of the Anaconda software package. I also learned the strength and weaknesses of Orange. Here is the youtube channel on Orange:

    4. \n
\n

https://www.youtube.com/channel/UClKKWBe2SCAEyv7ZNGhIe4g

\n
    \n
  1. I learned that the auto-machine learning software Accord.Net does its machine learning projects in C#. Because of this, I have decided that I am not going to utilize this software. Here is the link to that information though:
    2. \n
\n

https://github.com/accord-net/framework/wiki/Getting-started

\n
    \n
  1. I learned that I needed to learn tensorflow. I read half of the following book on it:
    2. \n
\n

https://www.amazon.com/Hands-Machine-Learning-Scikit-Learn-TensorFlow-ebook-dp-B07XGF2G87/dp/B07XGF2G87/ref=mt_other?_encoding=UTF8&me=&qid=

\n
    \n
  1. I learned how to install tensorflow in the anaconda package.
    2. \n
\n

https://towardsdatascience.com/https-medium-com-ekapope-v-install-tensorflow-and-keras-using-anaconda-navigator-without-command-line-b0bc41dbd038

\n
    \n
  1. To learn about Tensorflow, I watched the following Youtube video:
    2. \n
\n

https://www.youtube.com/watch?v=VwVg9jCtqaU

\n

Links associated with the above video:
\nhttps://codelabs.developers.google.com/codelabs/tensorflow-lab3-convolutions/index.html?index=..%2F..index#3

\n

https://colab.research.google.com/github/lmoroney/io19/blob/master/Zero%20to%20Hero/Rock-Paper-Scissors.ipynb

\n

https://www.tensorflow.org/tfx

\n
    \n
  1. I learned of the google developer codelabs websites. These websites are not just useful for learning Tensorflow, but they are good for learning many other languages and issues. They are the following:
    2. \n
\n

https://codelabs.developers.google.com/
\nhttps://codelabs.developers.google.com/learn-tensorflow/

\n
    \n
  1. I learned of the many models available on the tensorflow.org website. I learned that these models do not have python code embedded in them. I also learned that I needed to download these models as if they were a dataset.
    2. \n
\n

My discussion post on Kaggle when I learned of the above information:
\nhttps://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/discussion/181247

\n

The tensorflow hub where you can find models for images:
\nhttps://tfhub.dev/s?module-type=image-augmentation,image-classification,image-feature-vector,image-generator,image-object-detection,image-others,image-style-transfer,image-rnn-agent

\n
    \n
  1. I learned a little bit about TensorBoard
    2. \n
\n

https://www.tensorflow.org/tensorboard

\n
    \n
  1. I learned a little bit about the \"WhatIf\" model for tensorflow.
    2. \n
\n

https://pair-code.github.io/what-if-tool/

\n
    \n
  1. I learned that there was a certificate program (or test) to learn tensorflow. I plan on earning that certificate at some point in time.
    2. \n
\n

https://www.tensorflow.org/certificate

\n
    \n
  1. I learned of Pycharm and downloaded that program on my computer.
    2. \n
\n

https://www.jetbrains.com/pycharm/

\n
    \n
  1. I learned a little bit about Fast AI. In fact, I bought an O'Reilly book on it.
    2. \n
\n

https://www.amazon.com/Deep-Learning-Coders-fastai-PyTorch-ebook-dp-B08C2KM7NR/dp/B08C2KM7NR/ref=mt_other?_encoding=UTF8&me=&qid=1600926409

\n
    \n
  1. I learned a little bit about Hounsfield units, dicom files, and pydicom:
    2. \n
\n

https://pydicom.github.io/pydicom/dev/index.html
\nhttps://pydicom.github.io/pydicom/stable/index.html#
\nhttps://dicom.innolitics.com/ciods
\nhttps://pydicom.github.io/pydicom/0.9/working_with_pixel_data.html

\n
    \n
  1. I am learning about the kaggle API and determining if it is a good thing to use or not:
    2. \n
\n

https://github.com/Kaggle/kaggle-api

\n
    \n
  1. I also learned that the first-place team for another competition utilized many models to find the best solution.
    2. \n
\n

https://github.com/haqishen/SIIM-ISIC-Melanoma-Classification-1st-Place-Solution

\n
    \n
  1. I read Andrew Ng's book “Machine Learning Yearning.” Github found at following location:
    2. \n
\n

https://github.com/ajaymache/machine-learning-yearning

\n
    \n

  1. I learned a lot from others, but I also learned that I need to develop kernels of my own because other people’s kernel’s simply may not work, or they may have errors in them. In fact, I even recall copying a kernel and then running it and there were a lot of error codes which resulted after running them.

    2. \n

  2. I really like the following resource which gives python code examples which work with dicom files:

    3. \n
\n

https://www.programcreek.com/python/example/97517/dicom.read_file

\n

For above website I really like the last example because it converts the dicom file into an np array. My thinking was that if I convert the files into an np array that I could do

\n
    \n
  1. I also like the following resources because it teaches you how to use the models from the tensorflow hub:
    2. \n
\n

https://www.tensorflow.org/tutorials/images/transfer_learning_with_hub

\n

https://www.tensorflow.org/api_docs/python/tf/keras/utils/get_file

\n

Anyway, I hope that this post helps someone for this competition. I still have a lot to learn. Please let me know what your thoughts are if you have any.

\n

Thank you,
\nBrian

", "rawMarkdown": "To my fellow Kagglers like myself,\n\nThere are a lot of resources in this post which may be of interest to you. I hope you enjoy. Thank you.\n\nTo the creators of this competition,\n\nI want to thank you up front for the creation of this competition. I learned a lot as I attempted to do this OSIC Pulmonary Fibrosis competition. They are as follows:\n\n1. I got more comfortable with the Jupyter environment. \n\n2. I learned that combining kernels is not as simple as it sounds. I combined two kernels, ran it, and then had problems. I learned that code is dependent upon the code just preceding it.\n\n3. I learned about auto machine learning software. I did a free trial run of RapidMiner. This is a good software program. However, the software program that I have come to really enjoy was Orange. I like Orange because it is part of the Anaconda software package. I also learned the strength and weaknesses of Orange. Here is the youtube channel on Orange:\n\nhttps://www.youtube.com/channel/UClKKWBe2SCAEyv7ZNGhIe4g\n\n4. I learned that the auto-machine learning software Accord.Net does its machine learning projects in C#. Because of this, I have decided that I am not going to utilize this software. Here is the link to that information though:\n\nhttps://github.com/accord-net/framework/wiki/Getting-started\n\n5. I learned that I needed to learn tensorflow. I read half of the following book on it:\n\nhttps://www.amazon.com/Hands-Machine-Learning-Scikit-Learn-TensorFlow-ebook-dp-B07XGF2G87/dp/B07XGF2G87/ref=mt_other?_encoding=UTF8&me=&qid=\n\n6. I learned how to install tensorflow in the anaconda package.\n\nhttps://towardsdatascience.com/https-medium-com-ekapope-v-install-tensorflow-and-keras-using-anaconda-navigator-without-command-line-b0bc41dbd038\n\n7. To learn about Tensorflow, I watched the following Youtube video:\n\nhttps://www.youtube.com/watch?v=VwVg9jCtqaU\n\nLinks associated with the above video:\nhttps://codelabs.developers.google.com/codelabs/tensorflow-lab3-convolutions/index.html?index=..%2F..index#3\n\nhttps://colab.research.google.com/github/lmoroney/io19/blob/master/Zero%20to%20Hero/Rock-Paper-Scissors.ipynb\n\nhttps://www.tensorflow.org/tfx\n\n8. I learned of the google developer codelabs websites. These websites are not just useful for learning Tensorflow, but they are good for learning many other languages and issues. They are the following:\n\nhttps://codelabs.developers.google.com/\nhttps://codelabs.developers.google.com/learn-tensorflow/\n\n9. I learned of the many models available on the tensorflow.org website. I learned that these models do not have python code embedded in them. I also learned that I needed to download these models as if they were a dataset.\n\nMy discussion post on Kaggle when I learned of the above information:\nhttps://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/discussion/181247\n\nThe tensorflow hub where you can find models for images:\nhttps://tfhub.dev/s?module-type=image-augmentation,image-classification,image-feature-vector,image-generator,image-object-detection,image-others,image-style-transfer,image-rnn-agent\n\n10. I learned a little bit about TensorBoard\n\nhttps://www.tensorflow.org/tensorboard\n\n11. I learned a little bit about the \"WhatIf\" model for tensorflow.\n\nhttps://pair-code.github.io/what-if-tool/\n\n12. I learned that there was a certificate program (or test) to learn tensorflow. I plan on earning that certificate at some point in time.\n\nhttps://www.tensorflow.org/certificate\n\n13. I learned of Pycharm and downloaded that program on my computer.\n\nhttps://www.jetbrains.com/pycharm/\n\n14. I learned a little bit about Fast AI. In fact, I bought an O'Reilly book on it.\n\nhttps://www.amazon.com/Deep-Learning-Coders-fastai-PyTorch-ebook-dp-B08C2KM7NR/dp/B08C2KM7NR/ref=mt_other?_encoding=UTF8&me=&qid=1600926409\n\n15. I learned a little bit about Hounsfield units, dicom files, and pydicom:\n\nhttps://pydicom.github.io/pydicom/dev/index.html\nhttps://pydicom.github.io/pydicom/stable/index.html#\nhttps://dicom.innolitics.com/ciods\nhttps://pydicom.github.io/pydicom/0.9/working_with_pixel_data.html\n\n16. I am learning about the kaggle API and determining if it is a good thing to use or not:\n\nhttps://github.com/Kaggle/kaggle-api\n\n17. I also learned that the first-place team for another competition utilized many models to find the best solution.\n\nhttps://github.com/haqishen/SIIM-ISIC-Melanoma-Classification-1st-Place-Solution\n\n18. I read Andrew Ng's book “Machine Learning Yearning.” Github found at following location:\n\nhttps://github.com/ajaymache/machine-learning-yearning\n\n19. I learned a lot from others, but I also learned that I need to develop kernels of my own because other people’s kernel’s simply may not work, or they may have errors in them. In fact, I even recall copying a kernel and then running it and there were a lot of error codes which resulted after running them. \n\n20. I really like the following resource which gives python code examples which work with dicom files:\n\nhttps://www.programcreek.com/python/example/97517/dicom.read_file\n\nFor above website I really like the last example because it converts the dicom file into an np array. My thinking was that if I convert the files into an np array that I could do \n \n21. I also like the following resources because it teaches you how to use the models from the tensorflow hub:\n\nhttps://www.tensorflow.org/tutorials/images/transfer_learning_with_hub\n\nhttps://www.tensorflow.org/api_docs/python/tf/keras/utils/get_file\n\nAnyway, I hope that this post helps someone for this competition. I still have a lot to learn. Please let me know what your thoughts are if you have any. \n\nThank you,\nBrian", "votes": null }, { "id": "1024866", "postDate": "09/24/2020 07:18:23", "content": "

The numbering from my copy/paste function didn't survive the posting. There should be 21 items.

", "rawMarkdown": "The numbering from my copy/paste function didn't survive the posting. There should be 21 items.", "votes": null }, { "id": "1024956", "postDate": "09/24/2020 08:19:44", "content": "

Great to see that you learned a lot during the competition!
\nNext takeaway: you can always edit your original post ;)

", "rawMarkdown": "Great to see that you learned a lot during the competition!\nNext takeaway: you can always edit your original post ;)", "votes": null }, { "id": "1025078", "postDate": "09/24/2020 09:46:35", "content": "

I am aware of the feature. When I pressed the edit button my proper numbering appeared and then when I pressed \"post comment\" it went bad again.

", "rawMarkdown": "I am aware of the feature. When I pressed the edit button my proper numbering appeared and then when I pressed \"post comment\" it went bad again.", "votes": null }, { "id": "1025136", "postDate": "09/24/2020 10:56:09", "content": "

Good for you! I appreciate your determination to learn something new during this competition :)

", "rawMarkdown": "Good for you! I appreciate your determination to learn something new during this competition :)", "votes": null }, { "id": "1025193", "postDate": "09/24/2020 11:56:06", "content": "

Informative , Thanks for sharing @beamers

", "rawMarkdown": "Informative , Thanks for sharing @beamers", "votes": null }, { "id": "1025412", "postDate": "09/24/2020 14:35:22", "content": "

Very Much helpful and informative , Thanks for sharing

", "rawMarkdown": "Very Much helpful and informative , Thanks for sharing", "votes": null }, { "id": "1025418", "postDate": "09/24/2020 14:38:38", "content": "

Great learning

", "rawMarkdown": "Great learning", "votes": null }, { "id": "1025821", "postDate": "09/24/2020 19:29:23", "content": "

Really informative topic, thank you for collecting and share.

", "rawMarkdown": "Really informative topic, thank you for collecting and share.", "votes": null }, { "id": "1025909", "postDate": "09/24/2020 21:36:45", "content": "

Excellent resource examples and links !
\nI too. have learnt lots about dicom files.
\nThanks for posting.👍

", "rawMarkdown": "Excellent resource examples and links !\nI too. have learnt lots about dicom files.\nThanks for posting.👍", "votes": null }, { "id": "1026432", "postDate": "09/25/2020 10:14:51", "content": "

I learn't a lot too!

", "rawMarkdown": "I learn't a lot too!", "votes": null } ], "comments": [ { "id": 1024866, "author_name": "beamers", "author_url": "", "post_date": "09/24/2020 07:18:23", "content": "

The numbering from my copy/paste function didn't survive the posting. There should be 21 items.

", "votes": null, "replies": [ { "id": 1024956, "author_name": "ilu000", "author_url": "", "post_date": "09/24/2020 08:19:44", "content": "

Great to see that you learned a lot during the competition!
\nNext takeaway: you can always edit your original post ;)

", "votes": null, "replies": [] }, { "id": 1025078, "author_name": "beamers", "author_url": "", "post_date": "09/24/2020 09:46:35", "content": "

I am aware of the feature. When I pressed the edit button my proper numbering appeared and then when I pressed \"post comment\" it went bad again.

", "votes": null, "replies": [] } ] }, { "id": 1025136, "author_name": "aadhavvignesh", "author_url": "", "post_date": "09/24/2020 10:56:09", "content": "

Good for you! I appreciate your determination to learn something new during this competition :)

", "votes": null, "replies": [] }, { "id": 1025418, "author_name": "kurianbenoy", "author_url": "", "post_date": "09/24/2020 14:38:38", "content": "

Great learning

", "votes": null, "replies": [] }, { "id": 1025821, "author_name": "", "author_url": "", "post_date": "09/24/2020 19:29:23", "content": "

Really informative topic, thank you for collecting and share.

", "votes": null, "replies": [] }, { "id": 1025909, "author_name": "simo333", "author_url": "", "post_date": "09/24/2020 21:36:45", "content": "

Excellent resource examples and links !
\nI too. have learnt lots about dicom files.
\nThanks for posting.👍

", "votes": null, "replies": [] }, { "id": 1026432, "author_name": "geoffcc", "author_url": "", "post_date": "09/25/2020 10:14:51", "content": "

I learn't a lot too!

", "votes": null, "replies": [] }, { "id": 1025193, "author_name": "pinakimishrads", "author_url": "", "post_date": "09/24/2020 11:56:06", "content": "

Informative , Thanks for sharing @beamers

", "votes": null, "replies": [] }, { "id": 1025412, "author_name": "pinakimishrads", "author_url": "", "post_date": "09/24/2020 14:35:22", "content": "

Very Much helpful and informative , Thanks for sharing

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1024864": "To my fellow Kagglers like myself,\n\nThere are a lot of resources in this post which may be of interest to you. I hope you enjoy. Thank you.\n\nTo the creators of this competition,\n\nI want to thank you up front for the creation of this competition. I learned a lot as I attempted to do this OSIC Pulmonary Fibrosis competition. They are as follows:\n\n1. I got more comfortable with the Jupyter environment. \n\n2. I learned that combining kernels is not as simple as it sounds. I combined two kernels, ran it, and then had problems. I learned that code is dependent upon the code just preceding it.\n\n3. I learned about auto machine learning software. I did a free trial run of RapidMiner. This is a good software program. However, the software program that I have come to really enjoy was Orange. I like Orange because it is part of the Anaconda software package. I also learned the strength and weaknesses of Orange. Here is the youtube channel on Orange:\n\nhttps://www.youtube.com/channel/UClKKWBe2SCAEyv7ZNGhIe4g\n\n4. I learned that the auto-machine learning software Accord.Net does its machine learning projects in C#. Because of this, I have decided that I am not going to utilize this software. Here is the link to that information though:\n\nhttps://github.com/accord-net/framework/wiki/Getting-started\n\n5. I learned that I needed to learn tensorflow. I read half of the following book on it:\n\nhttps://www.amazon.com/Hands-Machine-Learning-Scikit-Learn-TensorFlow-ebook-dp-B07XGF2G87/dp/B07XGF2G87/ref=mt_other?_encoding=UTF8&me=&qid=\n\n6. I learned how to install tensorflow in the anaconda package.\n\nhttps://towardsdatascience.com/https-medium-com-ekapope-v-install-tensorflow-and-keras-using-anaconda-navigator-without-command-line-b0bc41dbd038\n\n7. To learn about Tensorflow, I watched the following Youtube video:\n\nhttps://www.youtube.com/watch?v=VwVg9jCtqaU\n\nLinks associated with the above video:\nhttps://codelabs.developers.google.com/codelabs/tensorflow-lab3-convolutions/index.html?index=..%2F..index#3\n\nhttps://colab.research.google.com/github/lmoroney/io19/blob/master/Zero%20to%20Hero/Rock-Paper-Scissors.ipynb\n\nhttps://www.tensorflow.org/tfx\n\n8. I learned of the google developer codelabs websites. These websites are not just useful for learning Tensorflow, but they are good for learning many other languages and issues. They are the following:\n\nhttps://codelabs.developers.google.com/\nhttps://codelabs.developers.google.com/learn-tensorflow/\n\n9. I learned of the many models available on the tensorflow.org website. I learned that these models do not have python code embedded in them. I also learned that I needed to download these models as if they were a dataset.\n\nMy discussion post on Kaggle when I learned of the above information:\nhttps://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/discussion/181247\n\nThe tensorflow hub where you can find models for images:\nhttps://tfhub.dev/s?module-type=image-augmentation,image-classification,image-feature-vector,image-generator,image-object-detection,image-others,image-style-transfer,image-rnn-agent\n\n10. I learned a little bit about TensorBoard\n\nhttps://www.tensorflow.org/tensorboard\n\n11. I learned a little bit about the \"WhatIf\" model for tensorflow.\n\nhttps://pair-code.github.io/what-if-tool/\n\n12. I learned that there was a certificate program (or test) to learn tensorflow. I plan on earning that certificate at some point in time.\n\nhttps://www.tensorflow.org/certificate\n\n13. I learned of Pycharm and downloaded that program on my computer.\n\nhttps://www.jetbrains.com/pycharm/\n\n14. I learned a little bit about Fast AI. In fact, I bought an O'Reilly book on it.\n\nhttps://www.amazon.com/Deep-Learning-Coders-fastai-PyTorch-ebook-dp-B08C2KM7NR/dp/B08C2KM7NR/ref=mt_other?_encoding=UTF8&me=&qid=1600926409\n\n15. I learned a little bit about Hounsfield units, dicom files, and pydicom:\n\nhttps://pydicom.github.io/pydicom/dev/index.html\nhttps://pydicom.github.io/pydicom/stable/index.html#\nhttps://dicom.innolitics.com/ciods\nhttps://pydicom.github.io/pydicom/0.9/working_with_pixel_data.html\n\n16. I am learning about the kaggle API and determining if it is a good thing to use or not:\n\nhttps://github.com/Kaggle/kaggle-api\n\n17. I also learned that the first-place team for another competition utilized many models to find the best solution.\n\nhttps://github.com/haqishen/SIIM-ISIC-Melanoma-Classification-1st-Place-Solution\n\n18. I read Andrew Ng's book “Machine Learning Yearning.” Github found at following location:\n\nhttps://github.com/ajaymache/machine-learning-yearning\n\n19. I learned a lot from others, but I also learned that I need to develop kernels of my own because other people’s kernel’s simply may not work, or they may have errors in them. In fact, I even recall copying a kernel and then running it and there were a lot of error codes which resulted after running them. \n\n20. I really like the following resource which gives python code examples which work with dicom files:\n\nhttps://www.programcreek.com/python/example/97517/dicom.read_file\n\nFor above website I really like the last example because it converts the dicom file into an np array. My thinking was that if I convert the files into an np array that I could do \n \n21. I also like the following resources because it teaches you how to use the models from the tensorflow hub:\n\nhttps://www.tensorflow.org/tutorials/images/transfer_learning_with_hub\n\nhttps://www.tensorflow.org/api_docs/python/tf/keras/utils/get_file\n\nAnyway, I hope that this post helps someone for this competition. I still have a lot to learn. Please let me know what your thoughts are if you have any. \n\nThank you,\nBrian", "1024866": "The numbering from my copy/paste function didn't survive the posting. There should be 21 items.", "1024956": "Great to see that you learned a lot during the competition!\nNext takeaway: you can always edit your original post ;)", "1025078": "I am aware of the feature. When I pressed the edit button my proper numbering appeared and then when I pressed \"post comment\" it went bad again.", "1025136": "Good for you! I appreciate your determination to learn something new during this competition :)", "1025193": "Informative , Thanks for sharing @beamers", "1025412": "Very Much helpful and informative , Thanks for sharing", "1025418": "Great learning", "1025821": "Really informative topic, thank you for collecting and share.", "1025909": "Excellent resource examples and links !\nI too. have learnt lots about dicom files.\nThanks for posting.👍", "1026432": "I learn't a lot too!" }, "source": "meta" }