{
  "id": 104536,
  "title": "Welcome from the organizers!",
  "url": "/competitions/understanding_cloud_organization/discussion/104536",
  "author_name": "",
  "post_date": "2019-08-17T15:06:40.885339100Z",
  "votes": 28,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Hi everyone! Welcome to the “Understanding Clouds from Satellite Images” competition.</p>\n\n<p>We are very pleased to share with you this dataset, which is the result of a collaboration of 70 scientists. At the beginning of this project was our curiosity about why clouds look the way they look. After looking at a lot of satellite pictures, we found some cloud structures that didn’t really conform to conventional cloud classes. We named them Sugar, Flower, Fish and Gravel. Then, to get more data on these cloud patterns, we organized a crowd-sourcing activity that yielded the labels that you will now use to build a machine learning model.</p>\n\n<p>Your models will help us extend our analysis, so that we can better understand what drives these clouds and what effects they have on our climate. Eventually, this will hopefully lead to better climate predictions. We are very excited to see which approaches the Kaggle community will take for this task. We also hope that we can learn a few tips and tricks on how to apply machine learning to satellite imagery in general.</p>\n\n<p>We will be checking the discussion board frequently to answer any questions you might have about the data. </p>\n\n<p>Best,\nStephan and Hauke</p>",
  "messages": [
    {
      "id": "601404",
      "postDate": "08/17/2019 15:06:40",
      "content": "<p>Hi everyone! Welcome to the “Understanding Clouds from Satellite Images” competition.</p>\n\n<p>We are very pleased to share with you this dataset, which is the result of a collaboration of 70 scientists. At the beginning of this project was our curiosity about why clouds look the way they look. After looking at a lot of satellite pictures, we found some cloud structures that didn’t really conform to conventional cloud classes. We named them Sugar, Flower, Fish and Gravel. Then, to get more data on these cloud patterns, we organized a crowd-sourcing activity that yielded the labels that you will now use to build a machine learning model.</p>\n\n<p>Your models will help us extend our analysis, so that we can better understand what drives these clouds and what effects they have on our climate. Eventually, this will hopefully lead to better climate predictions. We are very excited to see which approaches the Kaggle community will take for this task. We also hope that we can learn a few tips and tricks on how to apply machine learning to satellite imagery in general.</p>\n\n<p>We will be checking the discussion board frequently to answer any questions you might have about the data. </p>\n\n<p>Best,\nStephan and Hauke</p>",
      "rawMarkdown": "Hi everyone! Welcome to the “Understanding Clouds from Satellite Images” competition.\n\nWe are very pleased to share with you this dataset, which is the result of a collaboration of 70 scientists. At the beginning of this project was our curiosity about why clouds look the way they look. After looking at a lot of satellite pictures, we found some cloud structures that didn’t really conform to conventional cloud classes. We named them Sugar, Flower, Fish and Gravel. Then, to get more data on these cloud patterns, we organized a crowd-sourcing activity that yielded the labels that you will now use to build a machine learning model.\n\nYour models will help us extend our analysis, so that we can better understand what drives these clouds and what effects they have on our climate. Eventually, this will hopefully lead to better climate predictions. We are very excited to see which approaches the Kaggle community will take for this task. We also hope that we can learn a few tips and tricks on how to apply machine learning to satellite imagery in general.\n\nWe will be checking the discussion board frequently to answer any questions you might have about the data. \n\nBest,\nStephan and Hauke",
      "votes": null
    },
    {
      "id": "602344",
      "postDate": "08/19/2019 00:49:44",
      "content": "<p>Thanks for the organization. Interesting topic !</p>",
      "rawMarkdown": "Thanks for the organization. Interesting topic !",
      "votes": null
    },
    {
      "id": "606057",
      "postDate": "08/23/2019 06:18:32",
      "content": "<p>Thanks for the dataset which will help to understand more about satellite image</p>",
      "rawMarkdown": "Thanks for the dataset which will help to understand more about satellite image",
      "votes": null
    },
    {
      "id": "616314",
      "postDate": "09/03/2019 01:59:05",
      "content": "<p>Thanks for the organization. I am so excited.</p>",
      "rawMarkdown": "Thanks for the organization. I am so excited.",
      "votes": null
    },
    {
      "id": "629391",
      "postDate": "09/18/2019 18:05:16",
      "content": "<p>Hi guys,\ncan anyone help me to understand the encoded pixels in the train.csv</p>",
      "rawMarkdown": "Hi guys,\ncan anyone help me to understand the encoded pixels in the train.csv",
      "votes": null
    },
    {
      "id": "634155",
      "postDate": "09/26/2019 00:27:30",
      "content": "<p><a href=\"/raspstephan\">@raspstephan</a> May I ask is that possible for you to provide us with per-person annotations? I believe we can discover the pattern of annotation from each person (of course, by pure machine learning), then combine them. </p>",
      "rawMarkdown": "raspstephan May I ask is that possible for you to provide us with per-person annotations? I believe we can discover the pattern of annotation from each person (of course, by pure machine learning), then combine them.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 602344,
      "author_name": "grapestone5321",
      "author_url": "",
      "post_date": "08/19/2019 00:49:44",
      "content": "<p>Thanks for the organization. Interesting topic !</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 606057,
      "author_name": "roshanpanda",
      "author_url": "",
      "post_date": "08/23/2019 06:18:32",
      "content": "<p>Thanks for the dataset which will help to understand more about satellite image</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 616314,
      "author_name": "yyamamt29",
      "author_url": "",
      "post_date": "09/03/2019 01:59:05",
      "content": "<p>Thanks for the organization. I am so excited.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 629391,
      "author_name": "abhishekrock",
      "author_url": "",
      "post_date": "09/18/2019 18:05:16",
      "content": "<p>Hi guys,\ncan anyone help me to understand the encoded pixels in the train.csv</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 634155,
      "author_name": "khahuras",
      "author_url": "",
      "post_date": "09/26/2019 00:27:30",
      "content": "<p><a href=\"/raspstephan\">@raspstephan</a> May I ask is that possible for you to provide us with per-person annotations? I believe we can discover the pattern of annotation from each person (of course, by pure machine learning), then combine them. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "601404": "Hi everyone! Welcome to the “Understanding Clouds from Satellite Images” competition.\n\nWe are very pleased to share with you this dataset, which is the result of a collaboration of 70 scientists. At the beginning of this project was our curiosity about why clouds look the way they look. After looking at a lot of satellite pictures, we found some cloud structures that didn’t really conform to conventional cloud classes. We named them Sugar, Flower, Fish and Gravel. Then, to get more data on these cloud patterns, we organized a crowd-sourcing activity that yielded the labels that you will now use to build a machine learning model.\n\nYour models will help us extend our analysis, so that we can better understand what drives these clouds and what effects they have on our climate. Eventually, this will hopefully lead to better climate predictions. We are very excited to see which approaches the Kaggle community will take for this task. We also hope that we can learn a few tips and tricks on how to apply machine learning to satellite imagery in general.\n\nWe will be checking the discussion board frequently to answer any questions you might have about the data. \n\nBest,\nStephan and Hauke",
    "602344": "Thanks for the organization. Interesting topic !",
    "606057": "Thanks for the dataset which will help to understand more about satellite image",
    "616314": "Thanks for the organization. I am so excited.",
    "629391": "Hi guys,\ncan anyone help me to understand the encoded pixels in the train.csv",
    "634155": "raspstephan May I ask is that possible for you to provide us with per-person annotations? I believe we can discover the pattern of annotation from each person (of course, by pure machine learning), then combine them."
  },
  "source": "meta"
}