![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F116091%2Faf10da9cd0731f5a9233ff9fa4798687%2FKaggle_Upload.png?generation=1659508707531000&alt=media\")
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 Dr. Rachael Tatman has put together to learn a bit more about site etiquette, or Kaggle lingo.
\nRemember: 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.
\nHappy Modeling!
", "messages": [ { "id": 1784993, "postDate": "2022-05-11T17:11:39.300Z", "content": "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 Dr. Rachael Tatman has put together to learn a bit more about site etiquette, or Kaggle lingo.
\nRemember: 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.
\nHappy Modeling!
", "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 Dr. Rachael Tatman has put together to learn a bit more about [site etiquette](https://www.youtube.com/watch?v=aIus8si_Et0), or [Kaggle lingo](https://www.youtube.com/watch?v=sEJHyuWKd-s).\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).\n\nHappy Modeling!", "votes": 41 }, { "id": 1862224, "postDate": "2022-07-19T14:28:06.507Z", "content": "Hello, I'm weicheng. This is my first Kaggle competition. Let's get it!
", "rawMarkdown": "Hello, I'm weicheng. This is my first Kaggle competition. Let's get it!", "votes": 3, "replies": [ { "id": 1866375, "postDate": "2022-07-22T12:37:35.983Z", "rawMarkdown": "", "isDeleted": true } ] }, { "id": 1822201, "postDate": "2022-06-16T06:24:45.003Z", "content": "Hello!
\nI'm Dasha, this is my first very own competition I really going to participate (not just register to it and do nothing as usual…), so wish me luck haha
Greetings!
\nI am a Spanish guy with a lot of desire to learn more about machine learning and data analysis.
\nThis seems like a good place to take the basic knowledge I've gained so far further and keep improving.
\nI love this community and I am grateful to have the help of so many people here in Kaggle.
Kaggle seems complicated to me for some reason. Hope I'll learn to navigate the waters here.
", "rawMarkdown": "Kaggle seems complicated to me for some reason. Hope I'll learn to navigate the waters here.", "votes": 1 }, { "id": 1828100, "postDate": "2022-06-21T13:45:46.983Z", "content": "Hello there!
", "rawMarkdown": "Hello there!", "votes": 1, "replies": [ { "id": 1843901, "postDate": "2022-07-05T07:30:48.683Z", "content": "Hello there =))
", "rawMarkdown": "Hello there =)) " } ] }, { "id": 1810826, "postDate": "2022-06-04T00:02:38.430Z", "content": "Noice!
", "rawMarkdown": "Noice! \n", "votes": 1 }, { "id": 1805826, "postDate": "2022-05-30T14:34:29.347Z", "content": "Hi! I'm just learning machine learning
", "rawMarkdown": "Hi! I'm just learning machine learning", "votes": 1 }, { "id": 1805532, "postDate": "2022-05-30T09:15:12.113Z", "content": "Hi! New to Kaggle and data science. I'm happy to join this competition😊
", "rawMarkdown": "Hi! New to Kaggle and data science. I'm happy to join this competition😊", "votes": 1 }, { "id": 1798157, "postDate": "2022-05-22T18:07:40.200Z", "content": "hi, happy to be here, to learn and to be able to contribute.
", "rawMarkdown": "hi, happy to be here, to learn and to be able to contribute.", "votes": 1 }, { "id": 1820187, "postDate": "2022-06-14T12:04:58.180Z", "content": "Hi,
\nThis is Keerthi, MSc student in Data Analytics and this is my first kaggle competition
Hi Keethi,
\nThis is Ajay here. Completed my Masters in Data Analytics, few years ago, not used Kaggle much. This is my 2nd Kaggle Competition. Looking for team member to participate. Would you be interested?
hello! former scientist ready to get back into data science!
", "rawMarkdown": "hello! former scientist ready to get back into data science!", "votes": 2 }, { "id": 1801383, "postDate": "2022-05-25T17:05:46.437Z", "content": "This is my first real kaggle competition! I've had to use the platform for a class before and I've browsed it many times, but this challenge is up my alley, and I'm excited to get into it!
", "rawMarkdown": "This is my first real kaggle competition! I've had to use the platform for a class before and I've browsed it many times, but this challenge is up my alley, and I'm excited to get into it!", "votes": 2 }, { "id": 1798543, "postDate": "2022-05-23T06:04:33.300Z", "content": "Hi, I made account on kaggle few months back but started using it recently only… so yes Quite New to platform.
", "rawMarkdown": "Hi, I made account on kaggle few months back but started using it recently only... so yes Quite New to platform.", "votes": 2 }, { "id": 1785203, "postDate": "2022-05-11T23:22:10.533Z", "content": "Hi! Fairly new, excited for this one.
\nI noticed in the Getting Started notebook the authors said:
\n\n\nTo create a submission for this competition, we'll apply our model to the notebooks in the test set. Note that this is a Code Competition, which means that the test data we see here is only a small sample. When we submit our notebook for scoring, this example data will be replaced with the full test set of about 20,000 notebooks.
\n
Question: how do code submission competitions work? What in addition to the submission csv is provided for evaluation? How can I setup my notebook to make sure it runs correctly?
\nThanks!
", "rawMarkdown": "Hi! Fairly new, excited for this one.\n\nI noticed in the [Getting Started notebook](https://www.kaggle.com/code/ryanholbrook/getting-started-with-ai4code) the authors said:\n\n> To create a submission for this competition, we'll apply our model to the notebooks in the test set. Note that this is a Code Competition, which means that the test data we see here is only a small sample. When we submit our notebook for scoring, this example data will be replaced with the full test set of about 20,000 notebooks.\n\n\nQuestion: how do code submission competitions work? What in addition to the submission csv is provided for evaluation? How can I setup my notebook to make sure it runs correctly?\n\nThanks!\n", "votes": 2, "replies": [ { "id": 1785215, "postDate": "2022-05-11T23:52:03.780Z", "content": "To submit for a code competition, you have to create a notebook that then runs on the (unseen) test set. There's a sample test set when you create a new notebook (so you can check that your code runs), but the unseen set will have significantly more data when it runs.
\nIf you look at the \"Submission\" part of that notebook: https://www.kaggle.com/code/ryanholbrook/getting-started-with-ai4code#Submission you'll see they create a \"submission.csv\" file in the same format as the sample_submission.csv file - and that submission.csv file is what the auto checker will actually evaluate.
\nSo what I recommend is:
\nClick on the \"Code\" tab, then \"New Notebook\" - now you are creating a notebook for the competition (it's private by default; you can share publicly by clicking \"Share\" if you want).
Turn off the Internet toggle (code submissions aren't allowed to access the internet), and then write your code using the sample notebook as a guide to read from the test files and create a submission.csv
When you save your notebook it will run top to bottom against the sample test data, to make sure there are no runtime errors
Then, you can submit that saved notebook, either from the side panel in the notebook, or from the \"My Submissions\" page. It will run (might take awhile) and give you a score and put you on the leaderboard.
Hope that helps!
", "rawMarkdown": "To submit for a code competition, you have to create a notebook that then runs on the (unseen) test set. There's a sample test set when you create a new notebook (so you can check that your code runs), but the unseen set will have significantly more data when it runs.\n\nIf you look at the \"Submission\" part of that notebook: https://www.kaggle.com/code/ryanholbrook/getting-started-with-ai4code#Submission you'll see they create a \"submission.csv\" file in the same format as the sample_submission.csv file - and that submission.csv file is what the auto checker will actually evaluate.\n\nSo what I recommend is:\n\n- Click on the \"Code\" tab, then \"New Notebook\" - now you are creating a notebook for the competition (it's private by default; you can share publicly by clicking \"Share\" if you want).\n\n- Turn off the Internet toggle (code submissions aren't allowed to access the internet), and then write your code using the sample notebook as a guide to read from the test files and create a submission.csv\n\n- When you save your notebook it will run top to bottom against the sample test data, to make sure there are no runtime errors\n\n- Then, you can submit that saved notebook, either from the side panel in the notebook, or from the \"My Submissions\" page. It will run (might take awhile) and give you a score and put you on the leaderboard.\n\nHope that helps!", "votes": 7 }, { "id": 1785245, "postDate": "2022-05-12T00:26:49.420Z", "content": "Just adding to @chris62 's excellent comment: In this cell you see where we're using the trained model to create the test set predictions in model.predict(X_test). That model actually outputs an array of ranks for all the notebook cells, like [0, 2, 1, 3, 4, 5, ...]. If you create a model that also predicts ranks like that, you could just copy everything from that cell down into your own notebook and everything should \"just work\".
One way to get started would be to try to improve the feature engineering in that notebook. Scikit-learn has some other text processing options, but you could also explore some of the parameters of the TfidfTransformer (like, bi-grams might be useful), or try to create some features representing relationships between cells (some kind of text similarity feature, say).
", "rawMarkdown": "Just adding to @chris62 's excellent comment: In [this cell](https://admin.kaggle.com/code/ryanholbrook/getting-started-with-ai4code?scriptVersionId=95345057&cellId=53) you see where we're using the trained model to create the test set predictions in `model.predict(X_test)`. That model actually outputs an array of ranks for all the notebook cells, like `[0, 2, 1, 3, 4, 5, ...]`. If you create a model that also predicts ranks like that, you could just copy everything from that cell down into your own notebook and everything should \"just work\".\n\nOne way to get started would be to try to improve the feature engineering in that notebook. Scikit-learn has some other text processing options, but you could also explore some of the parameters of the TfidfTransformer (like, bi-grams might be useful), or try to create some features representing relationships between cells (some kind of text similarity feature, say).", "votes": 3 }, { "id": 1786387, "postDate": "2022-05-12T20:31:38.990Z", "content": "@chris62 Super helpful, thank you. Your comments helped me understand the reasoning behind the \"Save and Run\" dialog especially.
\n@ryanholbrook I appreciate the guidance and suggestions on where to get started.
Hi Ryan, I'm working on this competition. Can you suggest me how to improve my score.
\nThanks
hi just here for ml
", "rawMarkdown": "hi just here for ml" }, { "id": 2180998, "postDate": "2023-03-14T08:28:06.127Z", "content": "Hi all, I'm Frank. This is my first Kaggle competition. Let's dive in!
", "rawMarkdown": "Hi all, I'm Frank. This is my first Kaggle competition. Let's dive in!" }, { "id": 1978711, "postDate": "2022-10-09T00:02:03.227Z", "content": "Hello there!
", "rawMarkdown": "Hello there!" }, { "id": 1918656, "postDate": "2022-08-29T17:58:35.733Z", "content": "Hi, I am new here and just wanted to check what's happening. Kaggle seems like a great platform for ai-ml enthusiasts and I look forward to participating in future competitions. Peace.
", "rawMarkdown": "Hi, I am new here and just wanted to check what's happening. Kaggle seems like a great platform for ai-ml enthusiasts and I look forward to participating in future competitions. Peace." }, { "id": 1904927, "postDate": "2022-08-18T15:46:20.080Z", "content": "Hello, I am Cloe. This is my first time to learn code!
", "rawMarkdown": "Hello, I am Cloe. This is my first time to learn code!" }, { "id": 1904487, "postDate": "2022-08-18T08:12:58.770Z", "content": "Hi , I'm Anil , I'm new to competitions in kaggle.
", "rawMarkdown": "Hi , I'm Anil , I'm new to competitions in kaggle." }, { "id": 1900234, "postDate": "2022-08-15T20:13:51.543Z", "content": "hi new here.
", "rawMarkdown": "hi new here. \n" }, { "id": 1882246, "postDate": "2022-08-03T06:41:45.397Z", "content": "Hello everyone,
\nThis is Ajay from Bangalore, India. I haven't participated in Kaggle much. And don't have hackathon experience of NLP. Just know basic theoritical stuff before pre-deep Learning days. (Bag of Words, TF-IDF, Stemming, Lemmatizing). And have taken a course on Feedforward Neural Network so understand multiclass classification and Regression. Dont have practical coding experience with Deep Learning libraries (Tensorflow, Pytorch, Theano etc) and latest technologies (LSTM, BERT etc).
Given my limited background, will this competition be too difficult? pls advise.
\nI'm not able to place it - how to put it in a multiclass classification context. coz its more of a matching problem set A (is of comments/markdown) and set (B is of code). Looks more like finding a bipartite matching between two sets.
\n
And this above illustration was for just one notebook out of the 100000 in the Training data. How to scale it to the entire Training data consisting of Millions of Notebooks?
", "rawMarkdown": "Hello everyone,\nThis is Ajay from Bangalore, India. I haven't participated in Kaggle much. And don't have hackathon experience of NLP. Just know basic theoritical stuff before pre-deep Learning days. (Bag of Words, TF-IDF, Stemming, Lemmatizing). And have taken a course on Feedforward Neural Network so understand multiclass classification and Regression. Dont have practical coding experience with Deep Learning libraries (Tensorflow, Pytorch, Theano etc) and latest technologies (LSTM, BERT etc).\n\nGiven my limited background, will this competition be too difficult? pls advise.\n\nI'm not able to place it - how to put it in a multiclass classification context. coz its more of a matching problem set A (is of comments/markdown) and set (B is of code). Looks more like finding a bipartite matching between two sets.\n\n\n\n\nAnd this above illustration was for just **one notebook** out of the 100000 in the Training data. How to scale it to the entire Training data consisting of Millions of Notebooks?" }, { "id": 1881800, "postDate": "2022-08-02T19:39:52.317Z", "content": "Hi there! i have difficulties with deep learning but i hope i can get it
", "rawMarkdown": "Hi there! i have difficulties with deep learning but i hope i can get it" }, { "id": 1881081, "postDate": "2022-08-02T08:40:43.727Z", "content": "Great job!
", "rawMarkdown": "Great job!" }, { "id": 1880532, "postDate": "2022-08-01T19:42:23.570Z", "content": "Hello, Happy to participate
", "rawMarkdown": "Hello, Happy to participate" }, { "id": 1874035, "postDate": "2022-07-28T03:10:24.873Z", "content": "Me too! Nice to see you!
", "rawMarkdown": "Me too! Nice to see you!" }, { "id": 1871585, "postDate": "2022-07-26T10:52:47.517Z", "content": "Hi! I'm just learning machine learning.Nince to meet you!
", "rawMarkdown": "Hi! I'm just learning machine learning.Nince to meet you!" }, { "id": 1866504, "postDate": "2022-07-22T14:32:33.833Z", "content": "fresh man here!
", "rawMarkdown": "fresh man here!" }, { "id": 1861821, "postDate": "2022-07-19T09:11:04.783Z", "content": "Hi…👍💪
", "rawMarkdown": "Hi...👍💪 " }, { "id": 1861474, "postDate": "2022-07-19T04:00:58.893Z", "content": "Good luck guys.
", "rawMarkdown": "Good luck guys." }, { "id": 1849672, "postDate": "2022-07-09T17:44:18.237Z", "content": "I am trying to understand the files here. and below are my understandings. Please correct me if it is wrong
\nMy understanding about data and files:
Train Folder
\nTrain folder contains around 140,000 JSON files.(notebooks)
\nEach JSON file represents one notebook
\nEach JSON file contains two sections: cell_type and source
\nOrder of cells in both section is same.
\nCode cells are listed first
\nBelow code cells markdown cells are listed
\nCode cells are in correct order
\nMarkdown cells are not in correct order.
Train_Order.CSV
\nThis file contains correct order of cells for JSON files in train folder
train_ancestors.csv
\nIt is to track the copy of the notebooks
\nId = json file id or notebook id
\nAncestor_id: unique id to group json files (notebooks)copied from a common main file(main notebook)
\nParent_id: proper version of the notebook from which current notebook is copied
Test folder: around 20000 JSON files (notebooks)
", "rawMarkdown": "I am trying to understand the files here. and below are my understandings. Please correct me if it is wrong\nMy understanding about data and files:\n\nTrain Folder\nTrain folder contains around 140,000 JSON files.(notebooks)\nEach JSON file represents one notebook\nEach JSON file contains two sections: cell_type and source\nOrder of cells in both section is same.\nCode cells are listed first\nBelow code cells markdown cells are listed\nCode cells are in correct order\nMarkdown cells are not in correct order.\n\n\nTrain_Order.CSV\nThis file contains correct order of cells for JSON files in train folder\n\ntrain_ancestors.csv \nIt is to track the copy of the notebooks\nId = json file id or notebook id\nAncestor_id: unique id to group json files (notebooks)copied from a common main file(main notebook)\nParent_id: proper version of the notebook from which current notebook is copied\n\nTest folder: around 20000 JSON files (notebooks)\n" }, { "id": 1848989, "postDate": "2022-07-09T04:57:37.513Z", "content": "Hello, This is very helpful for me to build ML/AI skills.
", "rawMarkdown": "Hello, This is very helpful for me to build ML/AI skills." }, { "id": 1842895, "postDate": "2022-07-04T11:24:14.283Z", "content": "Thanks. I personally studied ML from datacamp and found it very useful.
", "rawMarkdown": "Thanks. I personally studied ML from datacamp and found it very useful." }, { "id": 1842538, "postDate": "2022-07-04T05:33:45.857Z", "content": "Hello, iam amazed by this platfrom.
", "rawMarkdown": "Hello, iam amazed by this platfrom." }, { "id": 1833732, "postDate": "2022-06-26T09:32:50.467Z", "content": "Hey everyone, I am here for a while but just activated my brain cells to do something 💪, Kaggle is amazing :)
", "rawMarkdown": "Hey everyone, I am here for a while but just activated my brain cells to do something 💪, Kaggle is amazing :)" }, { "id": 1832412, "postDate": "2022-06-25T02:59:54.557Z", "content": "Hello guys！
", "rawMarkdown": "Hello guys！" }, { "id": 1830525, "postDate": "2022-06-23T13:45:09.907Z", "content": "Hi all, great to join the competition!!
", "rawMarkdown": "Hi all, great to join the competition!!" }, { "id": 1828622, "postDate": "2022-06-21T22:13:56.370Z", "content": "Hey Everyone,
\nHope you are well. This is my first Kaggle competition and I'm excited to learn. I am currently having trouble understanding how to prepare the data and feed it into a model. I plan on building a simple learn to rank model using the tensor-flow ranking library. Any help would be appreciated. Thank you !
Hi. I am very new to Kaggle, but am really amazed by NLP. I did some classification based on Binomial classification, Multinomial classification, unsupervised features engineering as well as supervised features engineering using TF-IDF, LDA and NNMF. Can sb direct me to notebooks where I could use (transformers) word2vec, BERT, and the new T5 (is it available?).
", "rawMarkdown": "Hi. I am very new to Kaggle, but am really amazed by NLP. I did some classification based on Binomial classification, Multinomial classification, unsupervised features engineering as well as supervised features engineering using TF-IDF, LDA and NNMF. Can sb direct me to notebooks where I could use (transformers) word2vec, BERT, and the new T5 (is it available?)." }, { "id": 1795079, "postDate": "2022-05-19T11:55:26.163Z", "content": "I created a thread, my bad, but maybe I should have just asked my question here? 🤔 @addisonhoward, if you would be so kind and could please some light on the following, that would be greatly appreciated!!! Thank you so much! 🙏
", "rawMarkdown": "I created a thread, my bad, but maybe I should have just asked my question here? 🤔 @addisonhoward, if you would be so kind and could please some light on [the following](https://www.kaggle.com/competitions/AI4Code/discussion/325966), that would be greatly appreciated!!! Thank you so much! 🙏" }, { "id": 1791974, "postDate": "2022-05-16T14:06:50.770Z", "content": "New learner,like kaggle
", "rawMarkdown": "New learner,like kaggle " }, { "id": 1790978, "postDate": "2022-05-15T14:27:08.023Z", "content": "New to the site. I enjoy it so far.
", "rawMarkdown": "New to the site. I enjoy it so far." }, { "id": 1789993, "postDate": "2022-05-14T12:12:44.880Z", "content": "Hello! I am more or less a beginner - most of my machine learning/deep learning projects were just code copy pastes. But now I want to really get into machine learning and deep learning. And I think this competition is somewhat of a good starting point.
\nCouple questions about the challenge:
\n1) Is this a ML problem or DL problem? Sounds like DL to me.
\n2) What is my final output exactly supposed to be? Like when my model is tested, its supposed to sort the markdown and code cells?
\n3) How would I go about it? If my above assumption is true, My idea is to use the md cells explanations, and generate some code out of it and then compare it to the actual code cells. The one with the highest similarity would be paired with that markdown cell.
Hi! New to Kaggle and working on the Google Data Analytics Cert. I was wondering - where can you find tasks on a dataset?
", "rawMarkdown": "Hi! New to Kaggle and working on the Google Data Analytics Cert. I was wondering - where can you find tasks on a dataset?" }, { "id": 1786427, "postDate": "2022-05-12T22:09:28.787Z", "content": "Hi, 👋
\nNew here! Lot's of cool things to explore✌️
Hi … new to Data Science..
", "rawMarkdown": "Hi ... new to Data Science.. " }, { "id": 1785441, "postDate": "2022-05-12T05:24:52.047Z", "content": "Hi new to this site
\nQuestion: interested in NLP but have no clue where to start
https://www.kaggle.com/c/nlp-getting-started
\nStart from this challenge. This is for people who want to start with nlp
", "rawMarkdown": "https://www.kaggle.com/c/nlp-getting-started\n\nStart from this challenge. This is for people who want to start with nlp", "votes": 2 } ] }, { "id": 1908275, "postDate": "2022-08-21T13:52:41.683Z", "rawMarkdown": "", "isDeleted": true }, { "id": 1840448, "postDate": "2022-07-02T08:53:09.737Z", "rawMarkdown": "", "isDeleted": true }, { "id": 1833210, "postDate": "2022-06-25T18:21:28.380Z", "rawMarkdown": "", "isDeleted": true }, { "id": 1796414, "postDate": "2022-05-20T19:51:39.507Z", "rawMarkdown": "", "isDeleted": true }, { "id": 1793240, "postDate": "2022-05-17T17:52:54.023Z", "content": "Hi there, newbie to Kaggle as well. What do you suggest, going solo or looking for a team is best for a newcommer?
", "rawMarkdown": "Hi there, newbie to Kaggle as well. What do you suggest, going solo or looking for a team is best for a newcommer?", "isDeleted": true }, { "id": 1785545, "postDate": "2022-05-12T07:36:14.943Z", "rawMarkdown": "", "isDeleted": true }, { "id": 1838034, "postDate": "2022-06-30T06:35:49.453Z", "content": "This article really helped me thank you
", "rawMarkdown": "This article really helped me thank you" } ], "comments": [ { "id": 1862224, "author_name": "Hugo1995", "author_url": "", "post_date": "2022-07-19T14:28:06.507000", "content": "Hello, I'm weicheng. This is my first Kaggle competition. Let's get it!
", "votes": 3, "replies": [ { "id": 1866375, "author_name": "", "author_url": "", "post_date": "2022-07-22T12:37:35.983000", "content": "", "votes": 0, "replies": [] } ] }, { "id": 1822201, "author_name": "xufana", "author_url": "", "post_date": "2022-06-16T06:24:45.003000", "content": "Hello!
\nI'm Dasha, this is my first very own competition I really going to participate (not just register to it and do nothing as usual…), so wish me luck haha
Greetings!
\nI am a Spanish guy with a lot of desire to learn more about machine learning and data analysis.
\nThis seems like a good place to take the basic knowledge I've gained so far further and keep improving.
\nI love this community and I am grateful to have the help of so many people here in Kaggle.
Kaggle seems complicated to me for some reason. Hope I'll learn to navigate the waters here.
", "votes": 1, "replies": [] }, { "id": 1828100, "author_name": "Steve Le (Solokop)", "author_url": "", "post_date": "2022-06-21T13:45:46.983000", "content": "Hello there!
", "votes": 1, "replies": [ { "id": 1843901, "author_name": "DarkBubbleTea", "author_url": "", "post_date": "2022-07-05T07:30:48.683000", "content": "Hello there =))
", "votes": 0, "replies": [] } ] }, { "id": 1810826, "author_name": "Daniel", "author_url": "", "post_date": "2022-06-04T00:02:38.430000", "content": "Noice!
", "votes": 1, "replies": [] }, { "id": 1805826, "author_name": "Muhammad Fadhil Al-Amal", "author_url": "", "post_date": "2022-05-30T14:34:29.347000", "content": "Hi! I'm just learning machine learning
", "votes": 1, "replies": [] }, { "id": 1805532, "author_name": "Yumikiyo Osanai", "author_url": "", "post_date": "2022-05-30T09:15:12.113000", "content": "Hi! New to Kaggle and data science. I'm happy to join this competition😊
", "votes": 1, "replies": [] }, { "id": 1798157, "author_name": "Carlos Padilla Gomez", "author_url": "", "post_date": "2022-05-22T18:07:40.200000", "content": "hi, happy to be here, to learn and to be able to contribute.
", "votes": 1, "replies": [] }, { "id": 1820187, "author_name": "Keerthi Modupalli", "author_url": "", "post_date": "2022-06-14T12:04:58.180000", "content": "Hi,
\nThis is Keerthi, MSc student in Data Analytics and this is my first kaggle competition
Hi Keethi,
\nThis is Ajay here. Completed my Masters in Data Analytics, few years ago, not used Kaggle much. This is my 2nd Kaggle Competition. Looking for team member to participate. Would you be interested?
hello! former scientist ready to get back into data science!
", "votes": 2, "replies": [] }, { "id": 1801383, "author_name": "AlexShypula", "author_url": "", "post_date": "2022-05-25T17:05:46.437000", "content": "This is my first real kaggle competition! I've had to use the platform for a class before and I've browsed it many times, but this challenge is up my alley, and I'm excited to get into it!
", "votes": 2, "replies": [] }, { "id": 1798543, "author_name": "Dr Pradeep Jingar", "author_url": "", "post_date": "2022-05-23T06:04:33.300000", "content": "Hi, I made account on kaggle few months back but started using it recently only… so yes Quite New to platform.
", "votes": 2, "replies": [] }, { "id": 1785203, "author_name": "justinpwilliams", "author_url": "", "post_date": "2022-05-11T23:22:10.533000", "content": "Hi! Fairly new, excited for this one.
\nI noticed in the Getting Started notebook the authors said:
\n\n\nTo create a submission for this competition, we'll apply our model to the notebooks in the test set. Note that this is a Code Competition, which means that the test data we see here is only a small sample. When we submit our notebook for scoring, this example data will be replaced with the full test set of about 20,000 notebooks.
\n
Question: how do code submission competitions work? What in addition to the submission csv is provided for evaluation? How can I setup my notebook to make sure it runs correctly?
\nThanks!
", "votes": 2, "replies": [ { "id": 1785215, "author_name": "chris", "author_url": "", "post_date": "2022-05-11T23:52:03.780000", "content": "To submit for a code competition, you have to create a notebook that then runs on the (unseen) test set. There's a sample test set when you create a new notebook (so you can check that your code runs), but the unseen set will have significantly more data when it runs.
\nIf you look at the \"Submission\" part of that notebook: https://www.kaggle.com/code/ryanholbrook/getting-started-with-ai4code#Submission you'll see they create a \"submission.csv\" file in the same format as the sample_submission.csv file - and that submission.csv file is what the auto checker will actually evaluate.
\nSo what I recommend is:
\nClick on the \"Code\" tab, then \"New Notebook\" - now you are creating a notebook for the competition (it's private by default; you can share publicly by clicking \"Share\" if you want).
Turn off the Internet toggle (code submissions aren't allowed to access the internet), and then write your code using the sample notebook as a guide to read from the test files and create a submission.csv
When you save your notebook it will run top to bottom against the sample test data, to make sure there are no runtime errors
Then, you can submit that saved notebook, either from the side panel in the notebook, or from the \"My Submissions\" page. It will run (might take awhile) and give you a score and put you on the leaderboard.
Hope that helps!
", "votes": 7, "replies": [] }, { "id": 1785245, "author_name": "Ryan Holbrook", "author_url": "", "post_date": "2022-05-12T00:26:49.420000", "content": "Just adding to @chris62 's excellent comment: In this cell you see where we're using the trained model to create the test set predictions in model.predict(X_test). That model actually outputs an array of ranks for all the notebook cells, like [0, 2, 1, 3, 4, 5, ...]. If you create a model that also predicts ranks like that, you could just copy everything from that cell down into your own notebook and everything should \"just work\".
One way to get started would be to try to improve the feature engineering in that notebook. Scikit-learn has some other text processing options, but you could also explore some of the parameters of the TfidfTransformer (like, bi-grams might be useful), or try to create some features representing relationships between cells (some kind of text similarity feature, say).
", "votes": 3, "replies": [] }, { "id": 1786387, "author_name": "justinpwilliams", "author_url": "", "post_date": "2022-05-12T20:31:38.990000", "content": "@chris62 Super helpful, thank you. Your comments helped me understand the reasoning behind the \"Save and Run\" dialog especially.
\n@ryanholbrook I appreciate the guidance and suggestions on where to get started.
Hi Ryan, I'm working on this competition. Can you suggest me how to improve my score.
\nThanks
hi just here for ml
", "votes": 0, "replies": [] }, { "id": 2180998, "author_name": "Frank Mao", "author_url": "", "post_date": "2023-03-14T08:28:06.127000", "content": "Hi all, I'm Frank. This is my first Kaggle competition. Let's dive in!
", "votes": 0, "replies": [] }, { "id": 1978711, "author_name": "Eshwar 07", "author_url": "", "post_date": "2022-10-09T00:02:03.227000", "content": "Hello there!
", "votes": 0, "replies": [] }, { "id": 1918656, "author_name": "Rishikesh S", "author_url": "", "post_date": "2022-08-29T17:58:35.733000", "content": "Hi, I am new here and just wanted to check what's happening. Kaggle seems like a great platform for ai-ml enthusiasts and I look forward to participating in future competitions. Peace.
", "votes": 0, "replies": [] }, { "id": 1904927, "author_name": "LiuQinqin", "author_url": "", "post_date": "2022-08-18T15:46:20.080000", "content": "Hello, I am Cloe. This is my first time to learn code!
", "votes": 0, "replies": [] }, { "id": 1904487, "author_name": "Anil Red", "author_url": "", "post_date": "2022-08-18T08:12:58.770000", "content": "Hi , I'm Anil , I'm new to competitions in kaggle.
", "votes": 0, "replies": [] }, { "id": 1900234, "author_name": "mayankshrivastava", "author_url": "", "post_date": "2022-08-15T20:13:51.543000", "content": "hi new here.
", "votes": 0, "replies": [] }, { "id": 1882246, "author_name": "ajayram", "author_url": "", "post_date": "2022-08-03T06:41:45.397000", "content": "Hello everyone,
\nThis is Ajay from Bangalore, India. I haven't participated in Kaggle much. And don't have hackathon experience of NLP. Just know basic theoritical stuff before pre-deep Learning days. (Bag of Words, TF-IDF, Stemming, Lemmatizing). And have taken a course on Feedforward Neural Network so understand multiclass classification and Regression. Dont have practical coding experience with Deep Learning libraries (Tensorflow, Pytorch, Theano etc) and latest technologies (LSTM, BERT etc).
Given my limited background, will this competition be too difficult? pls advise.
\nI'm not able to place it - how to put it in a multiclass classification context. coz its more of a matching problem set A (is of comments/markdown) and set (B is of code). Looks more like finding a bipartite matching between two sets.
\n
And this above illustration was for just one notebook out of the 100000 in the Training data. How to scale it to the entire Training data consisting of Millions of Notebooks?
", "votes": 0, "replies": [] }, { "id": 1881800, "author_name": "Agustin222", "author_url": "", "post_date": "2022-08-02T19:39:52.317000", "content": "Hi there! i have difficulties with deep learning but i hope i can get it
", "votes": 0, "replies": [] }, { "id": 1881081, "author_name": "Elizabeth Xiu", "author_url": "", "post_date": "2022-08-02T08:40:43.727000", "content": "Great job!
", "votes": 0, "replies": [] }, { "id": 1880532, "author_name": "Mutuma", "author_url": "", "post_date": "2022-08-01T19:42:23.570000", "content": "Hello, Happy to participate
", "votes": 0, "replies": [] }, { "id": 1874035, "author_name": "Nguyen Thanh Trung", "author_url": "", "post_date": "2022-07-28T03:10:24.873000", "content": "Me too! Nice to see you!
", "votes": 0, "replies": [] }, { "id": 1871585, "author_name": "ruuu", "author_url": "", "post_date": "2022-07-26T10:52:47.517000", "content": "Hi! I'm just learning machine learning.Nince to meet you!
", "votes": 0, "replies": [] }, { "id": 1866504, "author_name": "Tori Wu", "author_url": "", "post_date": "2022-07-22T14:32:33.833000", "content": "fresh man here!
", "votes": 0, "replies": [] }, { "id": 1861821, "author_name": "Nguyen Thanh Trung", "author_url": "", "post_date": "2022-07-19T09:11:04.783000", "content": "Hi…👍💪
", "votes": 0, "replies": [] }, { "id": 1861474, "author_name": "", "author_url": "", "post_date": "2022-07-19T04:00:58.893000", "content": "", "votes": 0, "replies": [] }, { "id": 1849672, "author_name": "", "author_url": "", "post_date": "2022-07-09T17:44:18.237000", "content": "", "votes": 0, "replies": [] }, { "id": 1848989, "author_name": "", "author_url": "", "post_date": "2022-07-09T04:57:37.513000", "content": "", "votes": 0, "replies": [] }, { "id": 1842895, "author_name": "", "author_url": "", "post_date": "2022-07-04T11:24:14.283000", "content": "", "votes": 0, "replies": [] }, { "id": 1842538, "author_name": "", "author_url": "", "post_date": "2022-07-04T05:33:45.857000", "content": "", "votes": 0, "replies": [] }, { "id": 1833732, "author_name": "", "author_url": "", "post_date": "2022-06-26T09:32:50.467000", "content": "", "votes": 0, "replies": [] }, { "id": 1832412, "author_name": "", "author_url": "", "post_date": "2022-06-25T02:59:54.557000", "content": "", "votes": 0, "replies": [] }, { "id": 1830525, "author_name": "", "author_url": "", "post_date": "2022-06-23T13:45:09.907000", "content": "", "votes": 0, "replies": [] }, { "id": 1828622, "author_name": "", "author_url": "", "post_date": "2022-06-21T22:13:56.370000", "content": "", "votes": 0, "replies": [] }, { "id": 1828182, "author_name": "", "author_url": "", "post_date": "2022-06-21T14:27:02.497000", "content": "", "votes": 0, "replies": [] }, { "id": 1795079, "author_name": "", "author_url": "", "post_date": "2022-05-19T11:55:26.163000", "content": "", "votes": 0, "replies": [] }, { "id": 1791974, "author_name": "", "author_url": "", "post_date": "2022-05-16T14:06:50.770000", "content": "", "votes": 0, "replies": [] }, { "id": 1790978, "author_name": "", "author_url": "", "post_date": "2022-05-15T14:27:08.023000", "content": "", "votes": 0, "replies": [] }, { "id": 1789993, "author_name": "", "author_url": "", "post_date": "2022-05-14T12:12:44.880000", "content": "", "votes": 0, "replies": [] }, { "id": 1786506, "author_name": "", "author_url": "", "post_date": "2022-05-13T01:03:39.997000", "content": "", "votes": 0, "replies": [] }, { "id": 1786427, "author_name": "", "author_url": "", "post_date": "2022-05-12T22:09:28.787000", "content": "", "votes": 0, "replies": [] }, { "id": 1786173, "author_name": "", "author_url": "", "post_date": "2022-05-12T16:49:41.457000", "content": "", "votes": 0, "replies": [] }, { "id": 1785441, "author_name": "", "author_url": "", "post_date": "2022-05-12T05:24:52.047000", "content": "", "votes": 0, "replies": [ { "id": 1786650, "author_name": "", "author_url": "", "post_date": "2022-05-13T05:34:08.673000", "content": "", "votes": 2, "replies": [] } ] }, { "id": 1908275, "author_name": "", "author_url": "", "post_date": "2022-08-21T13:52:41.683000", "content": "", "votes": 0, "replies": [] }, { "id": 1840448, "author_name": "", "author_url": "", "post_date": "2022-07-02T08:53:09.737000", "content": "", "votes": 0, "replies": [] }, { "id": 1833210, "author_name": "", "author_url": "", "post_date": "2022-06-25T18:21:28.380000", "content": "", "votes": 0, "replies": [] }, { "id": 1796414, "author_name": "", "author_url": "", "post_date": "2022-05-20T19:51:39.507000", "content": "", "votes": 0, "replies": [] }, { "id": 1793240, "author_name": "", "author_url": "", "post_date": "2022-05-17T17:52:54.023000", "content": "", "votes": 0, "replies": [] }, { "id": 1785545, "author_name": "", "author_url": "", "post_date": "2022-05-12T07:36:14.943000", "content": "", "votes": 0, "replies": [] }, { "id": 1838034, "author_name": "", "author_url": "", "post_date": "2022-06-30T06:35:49.453000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "1784993": "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 Dr. Rachael Tatman has put together to learn a bit more about [site etiquette](https://www.youtube.com/watch?v=aIus8si_Et0), or [Kaggle lingo](https://www.youtube.com/watch?v=sEJHyuWKd-s).\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).\n\nHappy Modeling!", "1862224": "Hello, I'm weicheng. This is my first Kaggle competition. Let's get it!", "1822201": "Hello!\nI'm Dasha, this is my first very own competition I really going to participate (not just register to it and do nothing as usual...), so wish me luck haha", "1805848": "Greetings! \n\nI am a Spanish guy with a lot of desire to learn more about machine learning and data analysis.\nThis seems like a good place to take the basic knowledge I've gained so far further and keep improving.\nI love this community and I am grateful to have the help of so many people here in Kaggle.", "1904912": "Kaggle seems complicated to me for some reason. Hope I'll learn to navigate the waters here.", "1828100": "Hello there!", "1810826": "Noice! \n", "1805826": "Hi! I'm just learning machine learning", "1805532": "Hi! New to Kaggle and data science. I'm happy to join this competition😊", "1798157": "hi, happy to be here, to learn and to be able to contribute.", "1820187": "Hi,\nThis is Keerthi, MSc student in Data Analytics and this is my first kaggle competition", "1806252": "hello! former scientist ready to get back into data science!", "1801383": "This is my first real kaggle competition! I've had to use the platform for a class before and I've browsed it many times, but this challenge is up my alley, and I'm excited to get into it!", "1798543": "Hi, I made account on kaggle few months back but started using it recently only... so yes Quite New to platform.", "1785203": "Hi! Fairly new, excited for this one.\n\nI noticed in the [Getting Started notebook](https://www.kaggle.com/code/ryanholbrook/getting-started-with-ai4code) the authors said:\n\n> To create a submission for this competition, we'll apply our model to the notebooks in the test set. Note that this is a Code Competition, which means that the test data we see here is only a small sample. When we submit our notebook for scoring, this example data will be replaced with the full test set of about 20,000 notebooks.\n\n\nQuestion: how do code submission competitions work? What in addition to the submission csv is provided for evaluation? How can I setup my notebook to make sure it runs correctly?\n\nThanks!\n", "2394612": "hi just here for ml", "2180998": "Hi all, I'm Frank. This is my first Kaggle competition. Let's dive in!", "1978711": "Hello there!", "1918656": "Hi, I am new here and just wanted to check what's happening. Kaggle seems like a great platform for ai-ml enthusiasts and I look forward to participating in future competitions. Peace.", "1904927": "Hello, I am Cloe. This is my first time to learn code!", "1904487": "Hi , I'm Anil , I'm new to competitions in kaggle.", "1900234": "hi new here. \n", "1882246": "Hello everyone,\nThis is Ajay from Bangalore, India. I haven't participated in Kaggle much. And don't have hackathon experience of NLP. Just know basic theoritical stuff before pre-deep Learning days. (Bag of Words, TF-IDF, Stemming, Lemmatizing). And have taken a course on Feedforward Neural Network so understand multiclass classification and Regression. Dont have practical coding experience with Deep Learning libraries (Tensorflow, Pytorch, Theano etc) and latest technologies (LSTM, BERT etc).\n\nGiven my limited background, will this competition be too difficult? pls advise.\n\nI'm not able to place it - how to put it in a multiclass classification context. coz its more of a matching problem set A (is of comments/markdown) and set (B is of code). Looks more like finding a bipartite matching between two sets.\n\n\n\n\nAnd this above illustration was for just **one notebook** out of the 100000 in the Training data. How to scale it to the entire Training data consisting of Millions of Notebooks?", "1881800": "Hi there! i have difficulties with deep learning but i hope i can get it", "1881081": "Great job!", "1880532": "Hello, Happy to participate", "1874035": "Me too! Nice to see you!", "1871585": "Hi! I'm just learning machine learning.Nince to meet you!", "1866504": "fresh man here!", "1861821": "Hi...👍💪 ", "1861474": "Good luck guys.", "1849672": "I am trying to understand the files here. and below are my understandings. Please correct me if it is wrong\nMy understanding about data and files:\n\nTrain Folder\nTrain folder contains around 140,000 JSON files.(notebooks)\nEach JSON file represents one notebook\nEach JSON file contains two sections: cell_type and source\nOrder of cells in both section is same.\nCode cells are listed first\nBelow code cells markdown cells are listed\nCode cells are in correct order\nMarkdown cells are not in correct order.\n\n\nTrain_Order.CSV\nThis file contains correct order of cells for JSON files in train folder\n\ntrain_ancestors.csv \nIt is to track the copy of the notebooks\nId = json file id or notebook id\nAncestor_id: unique id to group json files (notebooks)copied from a common main file(main notebook)\nParent_id: proper version of the notebook from which current notebook is copied\n\nTest folder: around 20000 JSON files (notebooks)\n", "1848989": "Hello, This is very helpful for me to build ML/AI skills.", "1842895": "Thanks. I personally studied ML from datacamp and found it very useful.", "1842538": "Hello, iam amazed by this platfrom.", "1833732": "Hey everyone, I am here for a while but just activated my brain cells to do something 💪, Kaggle is amazing :)", "1832412": "Hello guys！", "1830525": "Hi all, great to join the competition!!", "1828622": "Hey Everyone, \nHope you are well. This is my first Kaggle competition and I'm excited to learn. I am currently having trouble understanding how to prepare the data and feed it into a model. I plan on building a simple learn to rank model using the tensor-flow ranking library. Any help would be appreciated. Thank you !", "1828182": "Hi. I am very new to Kaggle, but am really amazed by NLP. I did some classification based on Binomial classification, Multinomial classification, unsupervised features engineering as well as supervised features engineering using TF-IDF, LDA and NNMF. Can sb direct me to notebooks where I could use (transformers) word2vec, BERT, and the new T5 (is it available?).", "1795079": "I created a thread, my bad, but maybe I should have just asked my question here? 🤔 @addisonhoward, if you would be so kind and could please some light on [the following](https://www.kaggle.com/competitions/AI4Code/discussion/325966), that would be greatly appreciated!!! Thank you so much! 🙏", "1791974": "New learner,like kaggle ", "1790978": "New to the site. I enjoy it so far.", "1789993": "Hello! I am more or less a beginner - most of my machine learning/deep learning projects were just code copy pastes. But now I want to really get into machine learning and deep learning. And I think this competition is somewhat of a good starting point.\nCouple questions about the challenge:\n1) Is this a ML problem or DL problem? Sounds like DL to me.\n2) What is my final output exactly supposed to be? Like when my model is tested, its supposed to sort the markdown and code cells?\n3) How would I go about it? If my above assumption is true, My idea is to use the md cells explanations, and generate some code out of it and then compare it to the actual code cells. The one with the highest similarity would be paired with that markdown cell.", "1786506": "Hi! New to Kaggle and working on the Google Data Analytics Cert. I was wondering - where can you find tasks on a dataset?", "1786427": "Hi, 👋 \nNew here! Lot's of cool things to explore✌️", "1786173": "Hi ... new to Data Science.. ", "1785441": "Hi new to this site \nQuestion: interested in NLP but have no clue where to start", "1908275": "", "1840448": "", "1833210": "", "1796414": "", "1793240": "Hi there, newbie to Kaggle as well. What do you suggest, going solo or looking for a team is best for a newcommer?", "1785545": "", "1838034": "This article really helped me thank you" } }