{
  "id": 365899,
  "title": "What is the output we are expected to give?",
  "url": "/competitions/otto-recommender-system/discussion/365899",
  "author_name": "",
  "post_date": "2022-11-13T18:34:57.564109500Z",
  "votes": 3,
  "comment_count": 3,
  "views": 0,
  "content": "<p>The data section says<br>\n<strong>your task is to predict the next aid clicked after the session truncation, as well as the the remaining aids that are added to carts and orders; you may predict up to 20 values for each session type</strong></p>\n<p>This is not clear. By default there should be CLEAR examples showing what is expected. Its worse because the explanation on github doesn't match the one provided here</p>\n<p><strong>For each session id and type combination in the test set, you must predict the aid values in the label column, which is space delimited. You can predict up to 20 aid values per row. The file should contain a header and have the following format:</strong></p>\n<p>The first one is asking us to generate the next possible action/actions taken in the session and predict if any items will be added to cart/order. <br>\nThe second one is just asking us to predict the aid value OF THE GIVEN ENTRIES. These two are not the same. Which one are we expected to do in this competition?</p>",
  "messages": [
    {
      "id": "2028300",
      "postDate": "11/13/2022 18:34:57",
      "content": "<p>The data section says<br>\n<strong>your task is to predict the next aid clicked after the session truncation, as well as the the remaining aids that are added to carts and orders; you may predict up to 20 values for each session type</strong></p>\n<p>This is not clear. By default there should be CLEAR examples showing what is expected. Its worse because the explanation on github doesn't match the one provided here</p>\n<p><strong>For each session id and type combination in the test set, you must predict the aid values in the label column, which is space delimited. You can predict up to 20 aid values per row. The file should contain a header and have the following format:</strong></p>\n<p>The first one is asking us to generate the next possible action/actions taken in the session and predict if any items will be added to cart/order. <br>\nThe second one is just asking us to predict the aid value OF THE GIVEN ENTRIES. These two are not the same. Which one are we expected to do in this competition?</p>",
      "rawMarkdown": "The data section says\n**your task is to predict the next aid clicked after the session truncation, as well as the the remaining aids that are added to carts and orders; you may predict up to 20 values for each session type**\n\nThis is not clear. By default there should be CLEAR examples showing what is expected. Its worse because the explanation on github doesn't match the one provided here\n\n**For each session id and type combination in the test set, you must predict the aid values in the label column, which is space delimited. You can predict up to 20 aid values per row. The file should contain a header and have the following format:**\n\nThe first one is asking us to generate the next possible action/actions taken in the session and predict if any items will be added to cart/order. \nThe second one is just asking us to predict the aid value OF THE GIVEN ENTRIES. These two are not the same. Which one are we expected to do in this competition?",
      "votes": null
    },
    {
      "id": "2028726",
      "postDate": "11/14/2022 07:43:53",
      "content": "<p>Hi,<br>\nI think that they tell you the same thing. They are asking to predict the label column (that you don't have).<br>\nThe label should containt the aid of the next click if the row is a click row( you can make 20 guesses even though there is only 1 correct choice), for order and carts you have to guess up to 20 next cart/order actions.<br>\nThe second one says that you should give the values as space separated string, the first one focuses more on what you have to put inside it.</p>\n<p>I think the best would be to combine the two task description to have all the information in both places.</p>",
      "rawMarkdown": "Hi,\nI think that they tell you the same thing. They are asking to predict the label column (that you don't have).\nThe label should containt the aid of the next click if the row is a click row( you can make 20 guesses even though there is only 1 correct choice), for order and carts you have to guess up to 20 next cart/order actions.\nThe second one says that you should give the values as space separated string, the first one focuses more on what you have to put inside it.\n\nI think the best would be to combine the two task description to have all the information in both places.",
      "votes": null
    },
    {
      "id": "2084479",
      "postDate": "01/03/2023 14:30:03",
      "content": "<p><a href=\"https://www.kaggle.com/pietromaldini1\" target=\"_blank\">@pietromaldini1</a> what if there is only 1 correct choice case 1. but we have given 20 aid's which has the correct one.<br>\ncase2. the 20 aids given doesnt have the correct one.<br>\nin the earlier case will the score be high irrespective of 20 or not due to it contains correct one ?</p>",
      "rawMarkdown": "pietromaldini1 what if there is only 1 correct choice case 1. but we have given 20 aid's which has the correct one.\ncase2. the 20 aids given doesnt have the correct one.\nin the earlier case will the score be high irrespective of 20 or not due to it contains correct one ?",
      "votes": null
    },
    {
      "id": "2084531",
      "postDate": "01/03/2023 15:18:09",
      "content": "<p>In case 1, when you have the correct label among the 20 labels you predicted you will have a higher score.</p>",
      "rawMarkdown": "In case 1, when you have the correct label among the 20 labels you predicted you will have a higher score.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2028726,
      "author_name": "pietromaldini1",
      "author_url": "",
      "post_date": "11/14/2022 07:43:53",
      "content": "<p>Hi,<br>\nI think that they tell you the same thing. They are asking to predict the label column (that you don't have).<br>\nThe label should containt the aid of the next click if the row is a click row( you can make 20 guesses even though there is only 1 correct choice), for order and carts you have to guess up to 20 next cart/order actions.<br>\nThe second one says that you should give the values as space separated string, the first one focuses more on what you have to put inside it.</p>\n<p>I think the best would be to combine the two task description to have all the information in both places.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2084479,
          "author_name": "naiku007",
          "author_url": "",
          "post_date": "01/03/2023 14:30:03",
          "content": "<p><a href=\"https://www.kaggle.com/pietromaldini1\" target=\"_blank\">@pietromaldini1</a> what if there is only 1 correct choice case 1. but we have given 20 aid's which has the correct one.<br>\ncase2. the 20 aids given doesnt have the correct one.<br>\nin the earlier case will the score be high irrespective of 20 or not due to it contains correct one ?</p>",
          "votes": null,
          "replies": [
            {
              "id": 2084531,
              "author_name": "pietromaldini1",
              "author_url": "",
              "post_date": "01/03/2023 15:18:09",
              "content": "<p>In case 1, when you have the correct label among the 20 labels you predicted you will have a higher score.</p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2028300": "The data section says\n**your task is to predict the next aid clicked after the session truncation, as well as the the remaining aids that are added to carts and orders; you may predict up to 20 values for each session type**\n\nThis is not clear. By default there should be CLEAR examples showing what is expected. Its worse because the explanation on github doesn't match the one provided here\n\n**For each session id and type combination in the test set, you must predict the aid values in the label column, which is space delimited. You can predict up to 20 aid values per row. The file should contain a header and have the following format:**\n\nThe first one is asking us to generate the next possible action/actions taken in the session and predict if any items will be added to cart/order. \nThe second one is just asking us to predict the aid value OF THE GIVEN ENTRIES. These two are not the same. Which one are we expected to do in this competition?",
    "2028726": "Hi,\nI think that they tell you the same thing. They are asking to predict the label column (that you don't have).\nThe label should containt the aid of the next click if the row is a click row( you can make 20 guesses even though there is only 1 correct choice), for order and carts you have to guess up to 20 next cart/order actions.\nThe second one says that you should give the values as space separated string, the first one focuses more on what you have to put inside it.\n\nI think the best would be to combine the two task description to have all the information in both places.",
    "2084479": "pietromaldini1 what if there is only 1 correct choice case 1. but we have given 20 aid's which has the correct one.\ncase2. the 20 aids given doesnt have the correct one.\nin the earlier case will the score be high irrespective of 20 or not due to it contains correct one ?",
    "2084531": "In case 1, when you have the correct label among the 20 labels you predicted you will have a higher score."
  },
  "source": "meta"
}