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!
\nNew to Kaggle? Take a look at a few videos to learn a bit more about site etiquette, Kaggle lingo, and how to enter a competition using Kaggle Notebooks. Publish and share your models on Kaggle Models!
\nLooking for a team? Express your interest in joining a team through our Team Up feature.
\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.
", "messages": [ { "id": "3213782", "postDate": "05/30/2025 12:43:18", "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!
\nNew to Kaggle? Take a look at a few videos to learn a bit more about site etiquette, Kaggle lingo, and how to enter a competition using Kaggle Notebooks. Publish and share your models on Kaggle Models!
\nLooking for a team? Express your interest in joining a team through our Team Up feature.
\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.
", "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\n**New to Kaggle?** Take a look at a few videos 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). Publish and share your [models on Kaggle Models](https://www.kaggle.com/docs/models#publishing-a-model)!\n\n**Looking for a team?** Express your interest in joining a team through our [Team Up](https://www.kaggle.com/discussions/product-feedback/341195) feature.\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.", "votes": null }, { "id": "3215184", "postDate": "06/01/2025 18:21:56", "content": "Can I know which column I have to refer to see if an action is BFRB or non BFRB.
", "rawMarkdown": "Can I know which column I have to refer to see if an action is BFRB or non BFRB.", "votes": null }, { "id": "3215982", "postDate": "06/03/2025 00:58:52", "content": "I'm excited to participate in this competition and aim for a bronze medal.
\nWhat truly motivates me is the potential real-world applications that go far beyond the leaderboard.
The competition overview clearly describes the objective:
\n\"The goal of this competition is to build a model that distinguishes body-focused repetitive behaviors (BFRBs) from normal activities using sensor data collected from a wrist-worn device (Helios).\"
\nThis concept inspired me because it can be applied in various meaningful scenarios.
\nFor example, children with mental health conditions may exhibit behaviors like scratching their skin, biting their nails, or pulling their hair.
If we can monitor such behaviors quantitatively using the approach from this competition,
\nthen we could evaluate the effectiveness of different interventions or treatments over time.
\nThis would provide measurable insight into how well a certain strategy is helping a child improve.
I'm truly inspired by the combination of technical challenge and social impact,
\nand I look forward to learning from others and contributing to this community.
Let’s do our best!
", "rawMarkdown": "I'm excited to participate in this competition and aim for a bronze medal.\nWhat truly motivates me is the potential real-world applications that go far beyond the leaderboard.\n\nThe competition overview clearly describes the objective:\n\n\"The goal of this competition is to build a model that distinguishes body-focused repetitive behaviors (BFRBs) from normal activities using sensor data collected from a wrist-worn device (Helios).\"\n\nThis concept inspired me because it can be applied in various meaningful scenarios.\nFor example, children with mental health conditions may exhibit behaviors like scratching their skin, biting their nails, or pulling their hair.\n\nIf we can monitor such behaviors quantitatively using the approach from this competition,\nthen we could evaluate the effectiveness of different interventions or treatments over time.\nThis would provide measurable insight into how well a certain strategy is helping a child improve.\n\nI'm truly inspired by the combination of technical challenge and social impact,\nand I look forward to learning from others and contributing to this community.\n\nLet’s do our best!", "votes": null }, { "id": "3218149", "postDate": "06/05/2025 21:42:03", "content": "sequence_type
", "rawMarkdown": "sequence_type", "votes": null }, { "id": "3220636", "postDate": "06/09/2025 16:25:03", "content": "May want to mention that the behaviour column only appears in the training df
", "rawMarkdown": "May want to mention that the behaviour column only appears in the training df", "votes": null }, { "id": "3222290", "postDate": "06/12/2025 05:26:25", "content": "I am looking for someeone teaming up with :)
", "rawMarkdown": "I am looking for someeone teaming up with :)", "votes": null }, { "id": "3225597", "postDate": "06/16/2025 16:03:13", "content": "Thank you.
", "rawMarkdown": "Thank you.", "votes": null }, { "id": "3227191", "postDate": "06/18/2025 16:31:02", "content": "Hello all, This is Gpro here a beginner. Just documenting what I did.
\nMy first task was to wrap my head around the dataset. The Overview and data section were explanatory but, i wanted to see them physically. I was dumb enough to open it up in text editor but to see this message \"Data too large to load\". Then I uploaded it to my Google drive and accessed the data from google colab. I used Pandas to make a dataframe and found that train.csv is 1.5GB.
\nThen It took me a while to understand the structure of the data in \"train.csv\". Every row represents a snapshot of the instantaneous measurement that is recorded by the different sensors. Next on I wanted to visualize how a the accelerometer data looked like for a particular gesture. I was able to plot a line chart for \"acc_x\" for the \"Cheek - pinch skin\" gesture in the \"Seated Lean Non Dom - FACE DOWN\" position. (See chart 1. png).
\nI wanted to see if there is a visual pattern in all sensor data for this position and gesture. I plotted the first 20 of them (see chart 2.png).
So this was for the first 2 days..
\nNext I should be figuring out how to make my program learn the pattern from train.csv.
\nI am willing to learn. Suggestions are welcome..
Can we team up?
", "rawMarkdown": "Can we team up?", "votes": null }, { "id": "3228113", "postDate": "06/19/2025 17:40:11", "content": "Hi,
\nI understand that 'sequence_type' is the target column and the sensor data is the features, and training was done on the same using 'train' data. I also tried the 'test' data on trained model.
\nIn the submission notebook the function - 'def predict(sequence: pl.DataFrame, demographics: pl.DataFrame) -> str: '
\nis taking on two variables, I understand the input for 'sequence' part.
But I can't understand 'demographics' part. Should we use the 'train_demographics' data in anyway?
", "rawMarkdown": "Hi,\n\nI understand that 'sequence_type' is the target column and the sensor data is the features, and training was done on the same using 'train' data. I also tried the 'test' data on trained model. \n\nIn the submission notebook the function - 'def predict(sequence: pl.DataFrame, demographics: pl.DataFrame) -> str: '\nis taking on two variables, I understand the input for 'sequence' part.\n\nBut I can't understand 'demographics' part. Should we use the 'train_demographics' data in anyway?", "votes": null }, { "id": "3228132", "postDate": "06/19/2025 18:12:21", "content": "Demographics data is optional. You can use it in your training or leave it out
", "rawMarkdown": "Demographics data is optional. You can use it in your training or leave it out", "votes": null }, { "id": "3228990", "postDate": "06/20/2025 19:45:52", "content": "I don't really know how far the demographic dataset could contribute to the core model for this project, but here it is ! Some basic insights to start with. It may helps to further understand the bias in furture predictions. If you can find it useful for any type of interpretations in your current models, feel free to comment.
\n", "rawMarkdown": "I don't really know how far the demographic dataset could contribute to the core model for this project, but here it is ! Some basic insights to start with. It may helps to further understand the bias in furture predictions. If you can find it useful for any type of interpretations in your current models, feel free to comment.\n\n[demographic-dataset-eda](https://www.kaggle.com/code/pira245/cmi-demographic-dataset-eda-1)", "votes": null }, { "id": "3232416", "postDate": "06/25/2025 17:05:54", "content": "Hello!
\nAbout how to improve [Notebook Threw Exception]
\nIt runs fine on local test, but after submission I get Notebook Threw Exception. I am new to Kaggle and don't know how to resolve this. I would appreciate it if someone could tell me how to solve this problem or if I am writing the code incorrectly.
https://www.kaggle.com/code/naitoyuta/bfrb-ny-notebook
", "rawMarkdown": "Hello!\n\nAbout how to improve [Notebook Threw Exception]\nIt runs fine on local test, but after submission I get Notebook Threw Exception. I am new to Kaggle and don't know how to resolve this. I would appreciate it if someone could tell me how to solve this problem or if I am writing the code incorrectly.\n\nhttps://www.kaggle.com/code/naitoyuta/bfrb-ny-notebook", "votes": null }, { "id": "3235233", "postDate": "06/28/2025 23:26:34", "content": "Super welcoming to newbies! 😊 Thanks for the resources and reminders about site etiquette, competitions, and community guidelines. Excited to learn and grow with the Kaggle community! 🤓
", "rawMarkdown": "Super welcoming to newbies! 😊 Thanks for the resources and reminders about site etiquette, competitions, and community guidelines. Excited to learn and grow with the Kaggle community! 🤓", "votes": null }, { "id": "3245761", "postDate": "07/09/2025 20:07:17", "content": "Why is this a time series problem and not a sequential classification?
", "rawMarkdown": "Why is this a time series problem and not a sequential classification?", "votes": null }, { "id": "3252709", "postDate": "07/23/2025 08:32:28", "content": "Hello, it is a multi-class classification problem and not a timeseries problem
", "rawMarkdown": "Hello, it is a multi-class classification problem and not a timeseries problem", "votes": null }, { "id": "3255009", "postDate": "07/27/2025 18:13:35", "content": "Thank you Ravon.
", "rawMarkdown": "Thank you Ravon.", "votes": null }, { "id": "3266945", "postDate": "08/10/2025 02:56:38", "content": "Your current code does not appear to generate or save a file named submission.csv in /kaggle/working/. Once predictions are made, you must create this file explicitly and ensure it matches the format in sample_submission.csv
", "rawMarkdown": "Your current code does not appear to generate or save a file named submission.csv in /kaggle/working/. Once predictions are made, you must create this file explicitly and ensure it matches the format in sample_submission.csv", "votes": null }, { "id": "3274519", "postDate": "08/24/2025 23:42:42", "content": "Thanks for posting those useful resources! Really helpful for beginners.
", "rawMarkdown": "Thanks for posting those useful resources! Really helpful for beginners.", "votes": null } ], "comments": [ { "id": 3215184, "author_name": "akashsivakumar314", "author_url": "", "post_date": "06/01/2025 18:21:56", "content": "Can I know which column I have to refer to see if an action is BFRB or non BFRB.
", "votes": null, "replies": [ { "id": 3218149, "author_name": "nickfacciola", "author_url": "", "post_date": "06/05/2025 21:42:03", "content": "sequence_type
", "votes": null, "replies": [ { "id": 3225597, "author_name": "akashsivakumar314", "author_url": "", "post_date": "06/16/2025 16:03:13", "content": "Thank you.
", "votes": null, "replies": [] } ] } ] }, { "id": 3215982, "author_name": "toshiyukitanaka", "author_url": "", "post_date": "06/03/2025 00:58:52", "content": "I'm excited to participate in this competition and aim for a bronze medal.
\nWhat truly motivates me is the potential real-world applications that go far beyond the leaderboard.
The competition overview clearly describes the objective:
\n\"The goal of this competition is to build a model that distinguishes body-focused repetitive behaviors (BFRBs) from normal activities using sensor data collected from a wrist-worn device (Helios).\"
\nThis concept inspired me because it can be applied in various meaningful scenarios.
\nFor example, children with mental health conditions may exhibit behaviors like scratching their skin, biting their nails, or pulling their hair.
If we can monitor such behaviors quantitatively using the approach from this competition,
\nthen we could evaluate the effectiveness of different interventions or treatments over time.
\nThis would provide measurable insight into how well a certain strategy is helping a child improve.
I'm truly inspired by the combination of technical challenge and social impact,
\nand I look forward to learning from others and contributing to this community.
Let’s do our best!
", "votes": null, "replies": [ { "id": 3222290, "author_name": "ashmitachakraborty21", "author_url": "", "post_date": "06/12/2025 05:26:25", "content": "I am looking for someeone teaming up with :)
", "votes": null, "replies": [] } ] }, { "id": 3220636, "author_name": "isaiahmellace", "author_url": "", "post_date": "06/09/2025 16:25:03", "content": "May want to mention that the behaviour column only appears in the training df
", "votes": null, "replies": [] }, { "id": 3227191, "author_name": "gnanasenthilpro", "author_url": "", "post_date": "06/18/2025 16:31:02", "content": "Hello all, This is Gpro here a beginner. Just documenting what I did.
\nMy first task was to wrap my head around the dataset. The Overview and data section were explanatory but, i wanted to see them physically. I was dumb enough to open it up in text editor but to see this message \"Data too large to load\". Then I uploaded it to my Google drive and accessed the data from google colab. I used Pandas to make a dataframe and found that train.csv is 1.5GB.
\nThen It took me a while to understand the structure of the data in \"train.csv\". Every row represents a snapshot of the instantaneous measurement that is recorded by the different sensors. Next on I wanted to visualize how a the accelerometer data looked like for a particular gesture. I was able to plot a line chart for \"acc_x\" for the \"Cheek - pinch skin\" gesture in the \"Seated Lean Non Dom - FACE DOWN\" position. (See chart 1. png).
\nI wanted to see if there is a visual pattern in all sensor data for this position and gesture. I plotted the first 20 of them (see chart 2.png).
So this was for the first 2 days..
\nNext I should be figuring out how to make my program learn the pattern from train.csv.
\nI am willing to learn. Suggestions are welcome..
Can we team up?
", "votes": null, "replies": [] } ] }, { "id": 3228113, "author_name": "akashsivakumar314", "author_url": "", "post_date": "06/19/2025 17:40:11", "content": "Hi,
\nI understand that 'sequence_type' is the target column and the sensor data is the features, and training was done on the same using 'train' data. I also tried the 'test' data on trained model.
\nIn the submission notebook the function - 'def predict(sequence: pl.DataFrame, demographics: pl.DataFrame) -> str: '
\nis taking on two variables, I understand the input for 'sequence' part.
But I can't understand 'demographics' part. Should we use the 'train_demographics' data in anyway?
", "votes": null, "replies": [ { "id": 3228132, "author_name": "s617603", "author_url": "", "post_date": "06/19/2025 18:12:21", "content": "Demographics data is optional. You can use it in your training or leave it out
", "votes": null, "replies": [ { "id": 3255009, "author_name": "akashsivakumar314", "author_url": "", "post_date": "07/27/2025 18:13:35", "content": "Thank you Ravon.
", "votes": null, "replies": [] } ] } ] }, { "id": 3228990, "author_name": "pira245", "author_url": "", "post_date": "06/20/2025 19:45:52", "content": "I don't really know how far the demographic dataset could contribute to the core model for this project, but here it is ! Some basic insights to start with. It may helps to further understand the bias in furture predictions. If you can find it useful for any type of interpretations in your current models, feel free to comment.
\n", "votes": null, "replies": [] }, { "id": 3232416, "author_name": "naitoyuta", "author_url": "", "post_date": "06/25/2025 17:05:54", "content": "Hello!
\nAbout how to improve [Notebook Threw Exception]
\nIt runs fine on local test, but after submission I get Notebook Threw Exception. I am new to Kaggle and don't know how to resolve this. I would appreciate it if someone could tell me how to solve this problem or if I am writing the code incorrectly.
https://www.kaggle.com/code/naitoyuta/bfrb-ny-notebook
", "votes": null, "replies": [ { "id": 3266945, "author_name": "sunhithdontisunny", "author_url": "", "post_date": "08/10/2025 02:56:38", "content": "Your current code does not appear to generate or save a file named submission.csv in /kaggle/working/. Once predictions are made, you must create this file explicitly and ensure it matches the format in sample_submission.csv
", "votes": null, "replies": [] } ] }, { "id": 3235233, "author_name": "hamza1973", "author_url": "", "post_date": "06/28/2025 23:26:34", "content": "Super welcoming to newbies! 😊 Thanks for the resources and reminders about site etiquette, competitions, and community guidelines. Excited to learn and grow with the Kaggle community! 🤓
", "votes": null, "replies": [] }, { "id": 3245761, "author_name": "gauravsatwalkar", "author_url": "", "post_date": "07/09/2025 20:07:17", "content": "Why is this a time series problem and not a sequential classification?
", "votes": null, "replies": [ { "id": 3252709, "author_name": "yassermessahli", "author_url": "", "post_date": "07/23/2025 08:32:28", "content": "Hello, it is a multi-class classification problem and not a timeseries problem
", "votes": null, "replies": [] } ] }, { "id": 3274519, "author_name": "michaelfuzf64", "author_url": "", "post_date": "08/24/2025 23:42:42", "content": "Thanks for posting those useful resources! Really helpful for beginners.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "3213782": "**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\n**New to Kaggle?** Take a look at a few videos 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). Publish and share your [models on Kaggle Models](https://www.kaggle.com/docs/models#publishing-a-model)!\n\n**Looking for a team?** Express your interest in joining a team through our [Team Up](https://www.kaggle.com/discussions/product-feedback/341195) feature.\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.", "3215184": "Can I know which column I have to refer to see if an action is BFRB or non BFRB.", "3215982": "I'm excited to participate in this competition and aim for a bronze medal.\nWhat truly motivates me is the potential real-world applications that go far beyond the leaderboard.\n\nThe competition overview clearly describes the objective:\n\n\"The goal of this competition is to build a model that distinguishes body-focused repetitive behaviors (BFRBs) from normal activities using sensor data collected from a wrist-worn device (Helios).\"\n\nThis concept inspired me because it can be applied in various meaningful scenarios.\nFor example, children with mental health conditions may exhibit behaviors like scratching their skin, biting their nails, or pulling their hair.\n\nIf we can monitor such behaviors quantitatively using the approach from this competition,\nthen we could evaluate the effectiveness of different interventions or treatments over time.\nThis would provide measurable insight into how well a certain strategy is helping a child improve.\n\nI'm truly inspired by the combination of technical challenge and social impact,\nand I look forward to learning from others and contributing to this community.\n\nLet’s do our best!", "3218149": "sequence_type", "3220636": "May want to mention that the behaviour column only appears in the training df", "3222290": "I am looking for someeone teaming up with :)", "3225597": "Thank you.", "3227191": "Hello all, This is Gpro here a beginner. Just documenting what I did.\n\nMy first task was to wrap my head around the dataset. The Overview and data section were explanatory but, i wanted to see them physically. I was dumb enough to open it up in text editor but to see this message \"Data too large to load\". Then I uploaded it to my Google drive and accessed the data from google colab. I used Pandas to make a dataframe and found that train.csv is 1.5GB.\n\nThen It took me a while to understand the structure of the data in \"train.csv\". **Every row represents a snapshot of the instantaneous measurement that is recorded by the different sensors**. Next on I wanted to visualize how a the accelerometer data looked like for a particular gesture. I was able to plot a line chart for \"acc_x\" for the \"Cheek - pinch skin\" gesture in the \"Seated Lean Non Dom - FACE DOWN\" position. (See chart 1. png). \nI wanted to see if there is a visual pattern in all sensor data for this position and gesture. I plotted the first 20 of them (see chart 2.png).\n\nSo this was for the first 2 days..\n\nNext I should be figuring out how to make my program learn the pattern from train.csv. \nI am willing to learn. Suggestions are welcome..", "3227241": "Can we team up?", "3228113": "Hi,\n\nI understand that 'sequence_type' is the target column and the sensor data is the features, and training was done on the same using 'train' data. I also tried the 'test' data on trained model. \n\nIn the submission notebook the function - 'def predict(sequence: pl.DataFrame, demographics: pl.DataFrame) -> str: '\nis taking on two variables, I understand the input for 'sequence' part.\n\nBut I can't understand 'demographics' part. Should we use the 'train_demographics' data in anyway?", "3228132": "Demographics data is optional. You can use it in your training or leave it out", "3228990": "I don't really know how far the demographic dataset could contribute to the core model for this project, but here it is ! Some basic insights to start with. It may helps to further understand the bias in furture predictions. If you can find it useful for any type of interpretations in your current models, feel free to comment.\n\n[demographic-dataset-eda](https://www.kaggle.com/code/pira245/cmi-demographic-dataset-eda-1)", "3232416": "Hello!\n\nAbout how to improve [Notebook Threw Exception]\nIt runs fine on local test, but after submission I get Notebook Threw Exception. I am new to Kaggle and don't know how to resolve this. I would appreciate it if someone could tell me how to solve this problem or if I am writing the code incorrectly.\n\nhttps://www.kaggle.com/code/naitoyuta/bfrb-ny-notebook", "3235233": "Super welcoming to newbies! 😊 Thanks for the resources and reminders about site etiquette, competitions, and community guidelines. Excited to learn and grow with the Kaggle community! 🤓", "3245761": "Why is this a time series problem and not a sequential classification?", "3252709": "Hello, it is a multi-class classification problem and not a timeseries problem", "3255009": "Thank you Ravon.", "3266945": "Your current code does not appear to generate or save a file named submission.csv in /kaggle/working/. Once predictions are made, you must create this file explicitly and ensure it matches the format in sample_submission.csv", "3274519": "Thanks for posting those useful resources! Really helpful for beginners." }, "source": "meta" }