New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with!
\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!
\nNew to Kaggle? Take a look at a few videos we put together to learn a bit more about site etiquette, Kaggle lingo, and how to enter a competition using Kaggle Notebooks.
\nReady to dive into this competition? Review the Overview Description, ensure you understand the Competition Rules, and start to work with the Data!
\n\n", "messages": [ { "id": 919215, "postDate": "2020-07-07T19:02:08.167Z", "content": "Remember: Kaggle is for everyone. Whether you're teaming up or sharing tips in the competition forum, we expect everyone to follow our Kaggle community guidelines.
\n
New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with!
\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!
\nNew to Kaggle? Take a look at a few videos we put together to learn a bit more about site etiquette, Kaggle lingo, and how to enter a competition using Kaggle Notebooks.
\nReady to dive into this competition? Review the Overview Description, ensure you understand the Competition Rules, and start to work with the Data!
\n\n", "rawMarkdown": "New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with! \n\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!\n\nNew to Kaggle? Take a look at a few videos we put together to learn a bit more about [site etiquette](https://www.youtube.com/watch?v=aIus8si_Et0), [Kaggle lingo](https://www.youtube.com/watch?v=sEJHyuWKd-s), and [how to enter a competition using Kaggle Notebooks](https://www.youtube.com/watch?&v=GJBOMWpLpTQ).\n\nReady to dive into this competition? Review the [Overview Description](https://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/overview), ensure you understand the [Competition Rules](https://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/rules), and start to work with the [Data](https://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/data)!\n\n> **Remember**: Kaggle is for everyone. Whether you're teaming up or sharing tips in the competition forum, we expect everyone to follow our [Kaggle community guidelines](https://www.kaggle.com/community-guidelines).", "votes": 17 }, { "id": 971530, "postDate": "2020-08-15T16:30:05.053Z", "content": "Remember: Kaggle is for everyone. Whether you're teaming up or sharing tips in the competition forum, we expect everyone to follow our Kaggle community guidelines.
\n
Great! New to Kaggle as well. Do people usually build their codes on Kaggle notebook or somewhere else like Jupyter, and then copy it to Kaggle when they submit it?
", "rawMarkdown": "Great! New to Kaggle as well. Do people usually build their codes on Kaggle notebook or somewhere else like Jupyter, and then copy it to Kaggle when they submit it?", "votes": 1, "replies": [ { "id": 971702, "postDate": "2020-08-15T20:15:45.097Z", "content": "Wide variation. Often depends on the competition. Best to have basic hardware with a GPU to do local testing/development. Once you have something stable, load it in a Notebook. Kaggle Notebooks have GPU and TPU weekly quotas. Can also use Colab (free or paid) for additional resources.
\nMany probably just use Kaggle notebooks and you can definitely compete without your own hardware. Especially if a competition requires TPUs to run the models, you'll need the Notebooks or Colab.
\nI go back and forth regularly, depending on my quotas and how closely I want to interact with the model.
", "rawMarkdown": "Wide variation. Often depends on the competition. Best to have basic hardware with a GPU to do local testing/development. Once you have something stable, load it in a Notebook. Kaggle Notebooks have GPU and TPU weekly quotas. Can also use Colab (free or paid) for additional resources.\n\nMany probably just use Kaggle notebooks and you can definitely compete without your own hardware. Especially if a competition requires TPUs to run the models, you'll need the Notebooks or Colab.\n\nI go back and forth regularly, depending on my quotas and how closely I want to interact with the model.", "votes": 6 } ] }, { "id": 919649, "postDate": "2020-07-08T03:21:41.167Z", "content": "New on Kaggle
", "rawMarkdown": "New on Kaggle", "votes": 1, "replies": [ { "id": 924794, "postDate": "2020-07-11T16:22:05.720Z", "content": "Welcome! Hope you have a great competition.
", "rawMarkdown": "Welcome! Hope you have a great competition.", "votes": 1 } ] }, { "id": 944125, "postDate": "2020-07-24T21:06:06.610Z", "content": "Hi! \nWhat are minimal requirements for a laptop/desktop computer to perform decent models on this dataset?\nWould a laptop with i7, 8gb ram, 4gb graphics be any good to play around with dataset?
", "rawMarkdown": "Hi! \nWhat are minimal requirements for a laptop/desktop computer to perform decent models on this dataset?\nWould a laptop with i7, 8gb ram, 4gb graphics be any good to play around with dataset? ", "votes": 2, "replies": [ { "id": 949552, "postDate": "2020-07-28T17:53:52.070Z", "content": "I don't have much idea about the system specs requirements, but for this dataset (and the others present on Kaggle), Kaggle notebook would be a good choice since you can access its CPU and GPU processing capabilities via your browser. In that case, the minimum requirement would be a decent browser and a stable internet connection. I hope that answers your question.
", "rawMarkdown": "I don't have much idea about the system specs requirements, but for this dataset (and the others present on Kaggle), Kaggle notebook would be a good choice since you can access its CPU and GPU processing capabilities via your browser. In that case, the minimum requirement would be a decent browser and a stable internet connection. I hope that answers your question.", "votes": 2 }, { "id": 989540, "postDate": "2020-08-28T22:26:41.400Z", "content": "Yeah I agree with Yassh, it's probably worthwhile sticking to either the Kaggle GPU (remember to turn it off when you're done!) or something similar available online. GPUs for training models are very expensive, the decent ones are the NVIDIA RTX 2070 and better graphics cards but they're easily about 750USD+. But you don't necessarily need that at all to take part and even when you do you can rent one online for about 0.5USD per hour.
\nIn fact setting up my own GPU for machine learning was actually quite a pain in the neck and I have to be honest apart from the convenience and the quick response there isn't much benefit. As in, the actual training take a long time anyway (say a few seconds to a few hours depending on the dataset). These graphics cards even come with Machine Learning specific \"chips\" called tensor cores that perform the linear algebra underneath the hood very efficiently, but again, it's quite complicated and for beginners like us it's best to just get going first.
", "rawMarkdown": "Yeah I agree with Yassh, it's probably worthwhile sticking to either the Kaggle GPU (remember to turn it off when you're done!) or something similar available online. GPUs for training models are very expensive, the decent ones are the NVIDIA RTX 2070 and better graphics cards but they're easily about 750USD+. But you don't necessarily need that at all to take part and even when you do you can rent one online for about 0.5USD per hour.\n\nIn fact setting up my own GPU for machine learning was actually quite a pain in the neck and I have to be honest apart from the convenience and the quick response there isn't much benefit. As in, the actual training take a long time anyway (say a few seconds to a few hours depending on the dataset). These graphics cards even come with Machine Learning specific \"chips\" called tensor cores that perform the linear algebra underneath the hood very efficiently, but again, it's quite complicated and for beginners like us it's best to just get going first." } ] }, { "id": 1038333, "postDate": "2020-10-05T18:27:46.310Z", "content": "Hi, I'm new to Kaggle. I would like to know that, whether the chosen submissions are going to rerun for final submissions? I ask this because for some reason some times I get \"Submission Scoring Error\" and some times I don't while running the same code.
\nThanks in advance.
", "rawMarkdown": "Hi, I'm new to Kaggle. I would like to know that, whether the chosen submissions are going to rerun for final submissions? I ask this because for some reason some times I get \"Submission Scoring Error\" and some times I don't while running the same code.\n\nThanks in advance.", "replies": [ { "id": 1038366, "postDate": "2020-10-05T18:50:39.207Z", "content": "In this competition, the re-run is happening synchronously with your public submission. So if your submission returns a score on the public leaderboard, then you will be ranked on the private leaderboard. We will not be re-running the submissions again.
", "rawMarkdown": "In this competition, the re-run is happening synchronously with your public submission. So if your submission returns a score on the public leaderboard, then you will be ranked on the private leaderboard. We will not be re-running the submissions again." } ] }, { "id": 1017923, "postDate": "2020-09-19T10:47:58.597Z", "content": "can somebody tell me how to deal with missing FVC count of different weeks ?
", "rawMarkdown": "can somebody tell me how to deal with missing FVC count of different weeks ?\n" }, { "id": 1009814, "postDate": "2020-09-14T08:51:51.873Z", "content": "Hi, everyone
\nI have one question about this competition.
\nHow can I concatinate tabular data and image data?
Been registered with Kaggle a while and dipped in and out but now deciding to really get stuck into data science. Taken a few of the micro-courses and bought the Data Science from Scratch book by Joel Grus to help me on my journey.
", "rawMarkdown": "Been registered with Kaggle a while and dipped in and out but now deciding to really get stuck into data science. Taken a few of the micro-courses and bought the Data Science from Scratch book by Joel Grus to help me on my journey." }, { "id": 1000950, "postDate": "2020-09-07T00:12:09.890Z", "content": "Hi @juliaelliott, I am using a hybrid approach using both python and R, in my computer the .rmd script is working. Still figuring out in Kaggle; in the Rmarkdown language, I am not able to load several packages for the time being,
\n I am not sure about pointing/path of the python environment in kaggle, I have done in my personal laptop as use_condaenv(\"env\")(in that environment, I have installed all needed packages.)
Can you please provide a solution(or where can I find the solution) to this problem. If you can install the package in that environment it solves the problem. But I have used some other libraries as well all working in the competitions' python environment provided by Kaggle.
\nSince each time I have to run these codes to make the code works
\npy_install(\"scikit-learn\")
\npy_install(\"scipy\")
\npy_install(\"matplotlib\")
\npy_install(\"pydicom\")
\npy_install(\"scikit-image\")
\npy_install(\"scikit-image\")
Thanks in advance
", "rawMarkdown": "Hi @juliaelliott, I am using a hybrid approach using both python and R, in my computer the .rmd script is working. Still figuring out in Kaggle; in the Rmarkdown language, I am not able to load several packages for the time being,\n I am not sure about pointing/path of the python environment in kaggle, I have done in my personal laptop as use_condaenv(\"env\")(in that environment, I have installed all needed packages.)\n\nCan you please provide a solution(or where can I find the solution) to this problem. If you can install the package in that environment it solves the problem. But I have used some other libraries as well all working in the competitions' python environment provided by Kaggle.\nSince each time I have to run these codes to make the code works\npy_install(\"scikit-learn\")\npy_install(\"scipy\")\npy_install(\"matplotlib\")\npy_install(\"pydicom\")\npy_install(\"scikit-image\")\npy_install(\"scikit-image\")\n\nThanks in advance", "replies": [ { "id": 1003031, "postDate": "2020-09-08T16:00:45.517Z", "content": "@nishyanand - I'm not in a position to be providing troubleshooting advice. I recommend making a post that asks this question to peers in this competition.
", "rawMarkdown": "@nishyanand - I'm not in a position to be providing troubleshooting advice. I recommend making a post that asks this question to peers in this competition." } ] }, { "id": 994926, "postDate": "2020-09-02T04:30:23.500Z", "content": "New to kaggle as well.
", "rawMarkdown": "New to kaggle as well." }, { "id": 987075, "postDate": "2020-08-27T01:39:01.823Z", "content": "Hey guys! Could someone please explain what the Code Requirements mean?
\nIt states that there's a 4 hour runtime on gpu notebooks, so does that mean that we can only train for 4 hours?
The Code Requirements specify the systematic conditions applied to your submission notebook. So while you're welcome to create a notebook that runs for the 9 hours allowed by global notebook constraints, but if you try to submit with a notebook that exceeds the Code Requirements cap, then you will be prohibited from submitting. Many people competing in code competitions may opt to train offline or train in a separate notebook and export the trained model as an external data source into their submission notebook to run inference.
", "rawMarkdown": "The Code Requirements specify the systematic conditions applied to your submission notebook. So while you're welcome to create a notebook that runs for the 9 hours allowed by global notebook constraints, but if you try to submit with a notebook that exceeds the Code Requirements cap, then you will be prohibited from submitting. Many people competing in code competitions may opt to train offline or train in a separate notebook and export the trained model as an external data source into their submission notebook to run inference." }, { "id": 989528, "postDate": "2020-08-28T21:58:00.960Z", "content": "Does that mean that effectively there are no limits for the computation power you can use to train your models? As in the work doesn't necessarily need to be efficient?
", "rawMarkdown": "Does that mean that effectively there are no limits for the computation power you can use to train your models? As in the work doesn't necessarily need to be efficient?", "replies": [ { "id": 989531, "postDate": "2020-08-28T22:10:21.137Z", "content": "Yes. You can use any resources you want to train your model
", "rawMarkdown": "Yes. You can use any resources you want to train your model", "votes": 1 } ] } ] }, { "id": 975221, "postDate": "2020-08-18T07:49:57.790Z", "content": "Can anyone suggest notebooks for complete beginners with proper documentation of the code written.
\nOnly code seems quite difficult to understand in terms of understanding what is the approach
Hi
\nI'm new to kaggle and I have a concern. How can I improve my skills, I see other members notebooks and this sometimes make me feel that I would never get to their level of experience. For a new starter, what is the best practice to keep improving skills ?
", "rawMarkdown": "Hi\n\nI'm new to kaggle and I have a concern. How can I improve my skills, I see other members notebooks and this sometimes make me feel that I would never get to their level of experience. For a new starter, what is the best practice to keep improving skills ?", "replies": [ { "id": 989530, "postDate": "2020-08-28T22:00:51.963Z", "content": "I'm new too! I have been trying to just practise anything at all. I'm working on my coding skills (with challenges like advent of code) and going through the fastai course to help gain a better understanding of ML.
", "rawMarkdown": "I'm new too! I have been trying to just practise anything at all. I'm working on my coding skills (with challenges like advent of code) and going through the fastai course to help gain a better understanding of ML." } ] }, { "id": 971317, "postDate": "2020-08-15T11:38:44Z", "content": "Hi everyone,
\nit's also my first competition here - this topic sounds super interesting and I wish everybody good luck! 🤓
Hi everyone,
\nthis is my New challenge on Kaggle, good luck to everyone.
Hello! I'm new to code competitions. May I know can we use pre-trained models if the internet is disallowed? Because often times pre-trained models come from other sources, such as github, TFHub, etc.
", "rawMarkdown": "Hello! I'm new to code competitions. May I know can we use pre-trained models if the internet is disallowed? Because often times pre-trained models come from other sources, such as github, TFHub, etc.", "replies": [ { "id": 964992, "postDate": "2020-08-10T10:04:03.203Z", "content": "Hello @ardyharoen \nI hope this helps you.\nhttps://www.kaggle.com/paultimothymooney/how-to-use-pre-trained-models-in-kernels-on-kaggle
\n\nThank You
", "rawMarkdown": "Hello @ardyharoen \nI hope this helps you.\nhttps://www.kaggle.com/paultimothymooney/how-to-use-pre-trained-models-in-kernels-on-kaggle\n\nThank You", "votes": 2 }, { "id": 965025, "postDate": "2020-08-10T10:35:10.507Z", "content": "Thank you @palaksood97 ! Though I still don't understand.. in the notebook you gave, there was a section where the pre-trained model was downloaded from the internet. if we're not allowed to use the internet in this competition, how would we download the model? Thank you again!
", "rawMarkdown": "Thank you @palaksood97 ! Though I still don't understand.. in the notebook you gave, there was a section where the pre-trained model was downloaded from the internet. if we're not allowed to use the internet in this competition, how would we download the model? Thank you again!", "votes": 1 }, { "id": 965041, "postDate": "2020-08-10T10:48:26.880Z", "content": "Hello again,\nFirstly keras has some of the pretrained models such as ResNet50, VGG16, InceptionV3, etc..\nIf you wanna use these, then
\n\nfrom keras.applications.resnet50 import ResNet50 \nmodel = ResNet50()
\nmodel.summary()
else, \nI am not sure about this one, but one can download the model file externally, add it with the data, and finally use it.
\n\nDo let me know if the second one works for you.
\n\nThank you again
", "rawMarkdown": "Hello again,\nFirstly keras has some of the pretrained models such as ResNet50, VGG16, InceptionV3, etc..\nIf you wanna use these, then\n\nfrom keras.applications.resnet50 import ResNet50 \nmodel = ResNet50() \nmodel.summary() \n\nelse, \nI am not sure about this one, but one can download the model file externally, add it with the data, and finally use it.\n\nDo let me know if the second one works for you.\n\nThank you again", "votes": 3 }, { "id": 979838, "postDate": "2020-08-21T06:32:10.160Z", "content": "Thank you for the help!
", "rawMarkdown": "Thank you for the help!" }, { "id": 1002249, "postDate": "2020-09-08T01:24:05.117Z", "content": "\"May I know can we use pre-trained models\" -
\nTo answer your question, that depends on the competition. If I am not mistaken, yes, in this competition you can.
", "rawMarkdown": "\"May I know can we use pre-trained models\" -\n\nTo answer your question, that depends on the competition. If I am not mistaken, yes, in this competition you can." } ] }, { "id": 1013774, "postDate": "2020-09-17T01:10:17.910Z", "rawMarkdown": "", "isDeleted": true }, { "id": 1007911, "postDate": "2020-09-12T16:15:15.957Z", "rawMarkdown": "", "isDeleted": true }, { "id": 986471, "postDate": "2020-08-26T14:23:38.683Z", "rawMarkdown": "", "isDeleted": true, "replies": [ { "id": 989529, "postDate": "2020-08-28T21:59:29.810Z", "content": "Yeah I'm on roughly the same level myself. I'm using the fastai course to help myself get a better grasp of the topics as that API provides what is essentially a high level API for pytorch. I have found it really useful in enabling me to just get anything done at all.
", "rawMarkdown": "Yeah I'm on roughly the same level myself. I'm using the fastai course to help myself get a better grasp of the topics as that API provides what is essentially a high level API for pytorch. I have found it really useful in enabling me to just get anything done at all." }, { "id": 1002247, "postDate": "2020-09-08T01:19:11.307Z", "content": "If you have not already seen it, go to youtube and search for cs231n. It's a Stanford course, a series of lectures from the class. That might be something useful to you. Good luck!
", "rawMarkdown": "If you have not already seen it, go to youtube and search for cs231n. It's a Stanford course, a series of lectures from the class. That might be something useful to you. Good luck!" } ] }, { "id": 929535, "postDate": "2020-07-14T18:24:16.527Z", "rawMarkdown": "", "votes": 1, "isDeleted": true, "replies": [ { "id": 964979, "postDate": "2020-08-10T09:57:37.267Z", "content": "Welcome to this amazing community 😊
", "rawMarkdown": "Welcome to this amazing community 😊 ", "votes": 1 }, { "id": 971318, "postDate": "2020-08-15T11:39:16.497Z", "content": "Same here - have fun!
", "rawMarkdown": "Same here - have fun!" } ] } ], "comments": [ { "id": 971530, "author_name": "Jisoo Kim", "author_url": "", "post_date": "2020-08-15T16:30:05.053000", "content": "Great! New to Kaggle as well. Do people usually build their codes on Kaggle notebook or somewhere else like Jupyter, and then copy it to Kaggle when they submit it?
", "votes": 1, "replies": [ { "id": 971702, "author_name": "quadcore/Richard Epstein", "author_url": "", "post_date": "2020-08-15T20:15:45.097000", "content": "Wide variation. Often depends on the competition. Best to have basic hardware with a GPU to do local testing/development. Once you have something stable, load it in a Notebook. Kaggle Notebooks have GPU and TPU weekly quotas. Can also use Colab (free or paid) for additional resources.
\nMany probably just use Kaggle notebooks and you can definitely compete without your own hardware. Especially if a competition requires TPUs to run the models, you'll need the Notebooks or Colab.
\nI go back and forth regularly, depending on my quotas and how closely I want to interact with the model.
", "votes": 6, "replies": [] } ] }, { "id": 919649, "author_name": "Rahul Sawhney", "author_url": "", "post_date": "2020-07-08T03:21:41.167000", "content": "New on Kaggle
", "votes": 1, "replies": [ { "id": 924794, "author_name": "Gunjan Rawal", "author_url": "", "post_date": "2020-07-11T16:22:05.720000", "content": "Welcome! Hope you have a great competition.
", "votes": 1, "replies": [] } ] }, { "id": 944125, "author_name": "Shak", "author_url": "", "post_date": "2020-07-24T21:06:06.610000", "content": "Hi! \nWhat are minimal requirements for a laptop/desktop computer to perform decent models on this dataset?\nWould a laptop with i7, 8gb ram, 4gb graphics be any good to play around with dataset?
", "votes": 2, "replies": [ { "id": 949552, "author_name": "Yassh Ramchandani", "author_url": "", "post_date": "2020-07-28T17:53:52.070000", "content": "I don't have much idea about the system specs requirements, but for this dataset (and the others present on Kaggle), Kaggle notebook would be a good choice since you can access its CPU and GPU processing capabilities via your browser. In that case, the minimum requirement would be a decent browser and a stable internet connection. I hope that answers your question.
", "votes": 2, "replies": [] }, { "id": 989540, "author_name": "Arfan Araf", "author_url": "", "post_date": "2020-08-28T22:26:41.400000", "content": "Yeah I agree with Yassh, it's probably worthwhile sticking to either the Kaggle GPU (remember to turn it off when you're done!) or something similar available online. GPUs for training models are very expensive, the decent ones are the NVIDIA RTX 2070 and better graphics cards but they're easily about 750USD+. But you don't necessarily need that at all to take part and even when you do you can rent one online for about 0.5USD per hour.
\nIn fact setting up my own GPU for machine learning was actually quite a pain in the neck and I have to be honest apart from the convenience and the quick response there isn't much benefit. As in, the actual training take a long time anyway (say a few seconds to a few hours depending on the dataset). These graphics cards even come with Machine Learning specific \"chips\" called tensor cores that perform the linear algebra underneath the hood very efficiently, but again, it's quite complicated and for beginners like us it's best to just get going first.
", "votes": 0, "replies": [] } ] }, { "id": 1038333, "author_name": "Volkan", "author_url": "", "post_date": "2020-10-05T18:27:46.310000", "content": "Hi, I'm new to Kaggle. I would like to know that, whether the chosen submissions are going to rerun for final submissions? I ask this because for some reason some times I get \"Submission Scoring Error\" and some times I don't while running the same code.
\nThanks in advance.
", "votes": 0, "replies": [ { "id": 1038366, "author_name": "Julia Elliott", "author_url": "", "post_date": "2020-10-05T18:50:39.207000", "content": "In this competition, the re-run is happening synchronously with your public submission. So if your submission returns a score on the public leaderboard, then you will be ranked on the private leaderboard. We will not be re-running the submissions again.
", "votes": 0, "replies": [] } ] }, { "id": 1017923, "author_name": "VineetJAIN123", "author_url": "", "post_date": "2020-09-19T10:47:58.597000", "content": "can somebody tell me how to deal with missing FVC count of different weeks ?
", "votes": 0, "replies": [] }, { "id": 1009814, "author_name": "Taiki Ishii", "author_url": "", "post_date": "2020-09-14T08:51:51.873000", "content": "Hi, everyone
\nI have one question about this competition.
\nHow can I concatinate tabular data and image data?
Been registered with Kaggle a while and dipped in and out but now deciding to really get stuck into data science. Taken a few of the micro-courses and bought the Data Science from Scratch book by Joel Grus to help me on my journey.
", "votes": 0, "replies": [] }, { "id": 1000950, "author_name": "NishanthAnandanadarajah", "author_url": "", "post_date": "2020-09-07T00:12:09.890000", "content": "Hi @juliaelliott, I am using a hybrid approach using both python and R, in my computer the .rmd script is working. Still figuring out in Kaggle; in the Rmarkdown language, I am not able to load several packages for the time being,
\n I am not sure about pointing/path of the python environment in kaggle, I have done in my personal laptop as use_condaenv(\"env\")(in that environment, I have installed all needed packages.)
Can you please provide a solution(or where can I find the solution) to this problem. If you can install the package in that environment it solves the problem. But I have used some other libraries as well all working in the competitions' python environment provided by Kaggle.
\nSince each time I have to run these codes to make the code works
\npy_install(\"scikit-learn\")
\npy_install(\"scipy\")
\npy_install(\"matplotlib\")
\npy_install(\"pydicom\")
\npy_install(\"scikit-image\")
\npy_install(\"scikit-image\")
Thanks in advance
", "votes": 0, "replies": [ { "id": 1003031, "author_name": "Julia Elliott", "author_url": "", "post_date": "2020-09-08T16:00:45.517000", "content": "@nishyanand - I'm not in a position to be providing troubleshooting advice. I recommend making a post that asks this question to peers in this competition.
", "votes": 0, "replies": [] } ] }, { "id": 994926, "author_name": "Eric Olberding", "author_url": "", "post_date": "2020-09-02T04:30:23.500000", "content": "New to kaggle as well.
", "votes": 0, "replies": [] }, { "id": 987075, "author_name": "Mohammad Sharafat", "author_url": "", "post_date": "2020-08-27T01:39:01.823000", "content": "Hey guys! Could someone please explain what the Code Requirements mean?
\nIt states that there's a 4 hour runtime on gpu notebooks, so does that mean that we can only train for 4 hours?
The Code Requirements specify the systematic conditions applied to your submission notebook. So while you're welcome to create a notebook that runs for the 9 hours allowed by global notebook constraints, but if you try to submit with a notebook that exceeds the Code Requirements cap, then you will be prohibited from submitting. Many people competing in code competitions may opt to train offline or train in a separate notebook and export the trained model as an external data source into their submission notebook to run inference.
", "votes": 0, "replies": [] }, { "id": 989528, "author_name": "Arfan Araf", "author_url": "", "post_date": "2020-08-28T21:58:00.960000", "content": "Does that mean that effectively there are no limits for the computation power you can use to train your models? As in the work doesn't necessarily need to be efficient?
", "votes": 0, "replies": [ { "id": 989531, "author_name": "quadcore/Richard Epstein", "author_url": "", "post_date": "2020-08-28T22:10:21.137000", "content": "Yes. You can use any resources you want to train your model
", "votes": 1, "replies": [] } ] } ] }, { "id": 975221, "author_name": "Morpheus", "author_url": "", "post_date": "2020-08-18T07:49:57.790000", "content": "Can anyone suggest notebooks for complete beginners with proper documentation of the code written.
\nOnly code seems quite difficult to understand in terms of understanding what is the approach
Hi
\nI'm new to kaggle and I have a concern. How can I improve my skills, I see other members notebooks and this sometimes make me feel that I would never get to their level of experience. For a new starter, what is the best practice to keep improving skills ?
", "votes": 0, "replies": [ { "id": 989530, "author_name": "Arfan Araf", "author_url": "", "post_date": "2020-08-28T22:00:51.963000", "content": "I'm new too! I have been trying to just practise anything at all. I'm working on my coding skills (with challenges like advent of code) and going through the fastai course to help gain a better understanding of ML.
", "votes": 0, "replies": [] } ] }, { "id": 971317, "author_name": "Andreas Quauke", "author_url": "", "post_date": "2020-08-15T11:38:44", "content": "Hi everyone,
\nit's also my first competition here - this topic sounds super interesting and I wish everybody good luck! 🤓
Hi everyone,
\nthis is my New challenge on Kaggle, good luck to everyone.
Hello! I'm new to code competitions. May I know can we use pre-trained models if the internet is disallowed? Because often times pre-trained models come from other sources, such as github, TFHub, etc.
", "votes": 0, "replies": [ { "id": 964992, "author_name": "palak sood", "author_url": "", "post_date": "2020-08-10T10:04:03.203000", "content": "Hello @ardyharoen \nI hope this helps you.\nhttps://www.kaggle.com/paultimothymooney/how-to-use-pre-trained-models-in-kernels-on-kaggle
\n\nThank You
", "votes": 2, "replies": [] }, { "id": 965025, "author_name": "Ardy", "author_url": "", "post_date": "2020-08-10T10:35:10.507000", "content": "Thank you @palaksood97 ! Though I still don't understand.. in the notebook you gave, there was a section where the pre-trained model was downloaded from the internet. if we're not allowed to use the internet in this competition, how would we download the model? Thank you again!
", "votes": 1, "replies": [] }, { "id": 965041, "author_name": "palak sood", "author_url": "", "post_date": "2020-08-10T10:48:26.880000", "content": "Hello again,\nFirstly keras has some of the pretrained models such as ResNet50, VGG16, InceptionV3, etc..\nIf you wanna use these, then
\n\nfrom keras.applications.resnet50 import ResNet50 \nmodel = ResNet50()
\nmodel.summary()
else, \nI am not sure about this one, but one can download the model file externally, add it with the data, and finally use it.
\n\nDo let me know if the second one works for you.
\n\nThank you again
", "votes": 3, "replies": [] }, { "id": 979838, "author_name": "Ardy", "author_url": "", "post_date": "2020-08-21T06:32:10.160000", "content": "Thank you for the help!
", "votes": 0, "replies": [] }, { "id": 1002249, "author_name": "DM Grant", "author_url": "", "post_date": "2020-09-08T01:24:05.117000", "content": "\"May I know can we use pre-trained models\" -
\nTo answer your question, that depends on the competition. If I am not mistaken, yes, in this competition you can.
", "votes": 0, "replies": [] } ] }, { "id": 1013774, "author_name": "", "author_url": "", "post_date": "2020-09-17T01:10:17.910000", "content": "", "votes": 0, "replies": [] }, { "id": 1007911, "author_name": "", "author_url": "", "post_date": "2020-09-12T16:15:15.957000", "content": "", "votes": 0, "replies": [] }, { "id": 986471, "author_name": "", "author_url": "", "post_date": "2020-08-26T14:23:38.683000", "content": "", "votes": 0, "replies": [ { "id": 989529, "author_name": "Arfan Araf", "author_url": "", "post_date": "2020-08-28T21:59:29.810000", "content": "Yeah I'm on roughly the same level myself. I'm using the fastai course to help myself get a better grasp of the topics as that API provides what is essentially a high level API for pytorch. I have found it really useful in enabling me to just get anything done at all.
", "votes": 0, "replies": [] }, { "id": 1002247, "author_name": "DM Grant", "author_url": "", "post_date": "2020-09-08T01:19:11.307000", "content": "If you have not already seen it, go to youtube and search for cs231n. It's a Stanford course, a series of lectures from the class. That might be something useful to you. Good luck!
", "votes": 0, "replies": [] } ] }, { "id": 929535, "author_name": "", "author_url": "", "post_date": "2020-07-14T18:24:16.527000", "content": "", "votes": 1, "replies": [ { "id": 964979, "author_name": "palak sood", "author_url": "", "post_date": "2020-08-10T09:57:37.267000", "content": "Welcome to this amazing community 😊
", "votes": 1, "replies": [] }, { "id": 971318, "author_name": "Andreas Quauke", "author_url": "", "post_date": "2020-08-15T11:39:16.497000", "content": "Same here - have fun!
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "919215": "New to machine learning and data science? No question is too basic or too simple. Feel free to start your own thread, or use this thread as a place to post any first-timer clarifying questions for the Kaggle community to help you with! \n\nIf you would consider yourself a beginner but don't know where to get started, let other Kagglers help you take your first steps here!\n\nNew to Kaggle? Take a look at a few videos we put together to learn a bit more about [site etiquette](https://www.youtube.com/watch?v=aIus8si_Et0), [Kaggle lingo](https://www.youtube.com/watch?v=sEJHyuWKd-s), and [how to enter a competition using Kaggle Notebooks](https://www.youtube.com/watch?&v=GJBOMWpLpTQ).\n\nReady to dive into this competition? Review the [Overview Description](https://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/overview), ensure you understand the [Competition Rules](https://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/rules), and start to work with the [Data](https://www.kaggle.com/c/osic-pulmonary-fibrosis-progression/data)!\n\n> **Remember**: Kaggle is for everyone. Whether you're teaming up or sharing tips in the competition forum, we expect everyone to follow our [Kaggle community guidelines](https://www.kaggle.com/community-guidelines).", "971530": "Great! New to Kaggle as well. Do people usually build their codes on Kaggle notebook or somewhere else like Jupyter, and then copy it to Kaggle when they submit it?", "919649": "New on Kaggle", "944125": "Hi! \nWhat are minimal requirements for a laptop/desktop computer to perform decent models on this dataset?\nWould a laptop with i7, 8gb ram, 4gb graphics be any good to play around with dataset? ", "1038333": "Hi, I'm new to Kaggle. I would like to know that, whether the chosen submissions are going to rerun for final submissions? I ask this because for some reason some times I get \"Submission Scoring Error\" and some times I don't while running the same code.\n\nThanks in advance.", "1017923": "can somebody tell me how to deal with missing FVC count of different weeks ?\n", "1009814": "Hi, everyone\nI have one question about this competition.\nHow can I concatinate tabular data and image data?", "1002622": "Been registered with Kaggle a while and dipped in and out but now deciding to really get stuck into data science. Taken a few of the micro-courses and bought the Data Science from Scratch book by Joel Grus to help me on my journey.", "1000950": "Hi @juliaelliott, I am using a hybrid approach using both python and R, in my computer the .rmd script is working. Still figuring out in Kaggle; in the Rmarkdown language, I am not able to load several packages for the time being,\n I am not sure about pointing/path of the python environment in kaggle, I have done in my personal laptop as use_condaenv(\"env\")(in that environment, I have installed all needed packages.)\n\nCan you please provide a solution(or where can I find the solution) to this problem. If you can install the package in that environment it solves the problem. But I have used some other libraries as well all working in the competitions' python environment provided by Kaggle.\nSince each time I have to run these codes to make the code works\npy_install(\"scikit-learn\")\npy_install(\"scipy\")\npy_install(\"matplotlib\")\npy_install(\"pydicom\")\npy_install(\"scikit-image\")\npy_install(\"scikit-image\")\n\nThanks in advance", "994926": "New to kaggle as well.", "987075": "Hey guys! Could someone please explain what the Code Requirements mean?\nIt states that there's a 4 hour runtime on gpu notebooks, so does that mean that we can only train for 4 hours?", "975221": "Can anyone suggest notebooks for complete beginners with proper documentation of the code written.\nOnly code seems quite difficult to understand in terms of understanding what is the approach ", "974318": "Hi\n\nI'm new to kaggle and I have a concern. How can I improve my skills, I see other members notebooks and this sometimes make me feel that I would never get to their level of experience. For a new starter, what is the best practice to keep improving skills ?", "971317": "Hi everyone, \nit's also my first competition here - this topic sounds super interesting and I wish everybody good luck! 🤓", "969685": "Hi everyone,\nthis is my New challenge on Kaggle, good luck to everyone.", "963446": "Hello! I'm new to code competitions. May I know can we use pre-trained models if the internet is disallowed? Because often times pre-trained models come from other sources, such as github, TFHub, etc.", "1013774": "", "1007911": "", "986471": "", "929535": "" } }