{
  "id": 62085,
  "title": "A simple question about 'PredictionString'",
  "url": "/competitions/google-ai-open-images-visual-relationship-track/discussion/62085",
  "author_name": "",
  "post_date": "2018-07-27T10:12:35.859983700Z",
  "votes": 3,
  "comment_count": 2,
  "views": 0,
  "content": "<p>maybe sorry, I'm newer.</p>\n\n<p>I noticed the 'PredictionString' format of VRD_sample_submission.csv</p>\n\n<p>I can understand the PredictionString  format like:</p>\n\n<pre><code>[LabelName1,XMin1,XMax1,YMin1,YMax1,LabelName2(or attribute),XMin2,XMax2,YMin2,YMax2,RelationLabel]*\n</code></pre>\n\n<p>but in the VRD_sample_submission.csv is like:</p>\n\n<pre><code>[float,LabelName1,XMin1,XMax1,YMin1,YMax1,LabelName2(or attribute),XMin2,XMax2,YMin2,YMax2,RelationLabel]*\n</code></pre>\n\n<p>How should I understand the first float value in each tuple, is the probability of each tuple of relationships?</p>",
  "messages": [
    {
      "id": "362873",
      "postDate": "07/27/2018 10:12:35",
      "content": "<p>maybe sorry, I'm newer.</p>\n\n<p>I noticed the 'PredictionString' format of VRD_sample_submission.csv</p>\n\n<p>I can understand the PredictionString  format like:</p>\n\n<pre><code>[LabelName1,XMin1,XMax1,YMin1,YMax1,LabelName2(or attribute),XMin2,XMax2,YMin2,YMax2,RelationLabel]*\n</code></pre>\n\n<p>but in the VRD_sample_submission.csv is like:</p>\n\n<pre><code>[float,LabelName1,XMin1,XMax1,YMin1,YMax1,LabelName2(or attribute),XMin2,XMax2,YMin2,YMax2,RelationLabel]*\n</code></pre>\n\n<p>How should I understand the first float value in each tuple, is the probability of each tuple of relationships?</p>",
      "rawMarkdown": "maybe sorry, I'm newer.\n\nI noticed the 'PredictionString' format of VRD_sample_submission.csv\n\nI can understand the PredictionString  format like:\n        \n\n    [LabelName1,XMin1,XMax1,YMin1,YMax1,LabelName2(or attribute),XMin2,XMax2,YMin2,YMax2,RelationLabel]*\n\nbut in the VRD_sample_submission.csv is like:\n\n    [float,LabelName1,XMin1,XMax1,YMin1,YMax1,LabelName2(or attribute),XMin2,XMax2,YMin2,YMax2,RelationLabel]*\n\nHow should I understand the first float value in each tuple, is the probability of each tuple of relationships?",
      "votes": null
    },
    {
      "id": "362979",
      "postDate": "07/27/2018 14:17:07",
      "content": "<p>The first float is the confidence score -&gt; how much your model is convinced that this particular prediction is correct (take a look at the metrics use in this competition - they make use of this information).</p>",
      "rawMarkdown": "The first float is the confidence score -&gt; how much your model is convinced that this particular prediction is correct (take a look at the metrics use in this competition - they make use of this information).",
      "votes": null
    },
    {
      "id": "362981",
      "postDate": "07/27/2018 14:21:23",
      "content": "<p>thanks a lot.</p>",
      "rawMarkdown": "thanks a lot.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 362979,
      "author_name": "radek1",
      "author_url": "",
      "post_date": "07/27/2018 14:17:07",
      "content": "<p>The first float is the confidence score -&gt; how much your model is convinced that this particular prediction is correct (take a look at the metrics use in this competition - they make use of this information).</p>",
      "votes": null,
      "replies": [
        {
          "id": 362981,
          "author_name": "xffjlg",
          "author_url": "",
          "post_date": "07/27/2018 14:21:23",
          "content": "<p>thanks a lot.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "362873": "maybe sorry, I'm newer.\n\nI noticed the 'PredictionString' format of VRD_sample_submission.csv\n\nI can understand the PredictionString  format like:\n        \n\n    [LabelName1,XMin1,XMax1,YMin1,YMax1,LabelName2(or attribute),XMin2,XMax2,YMin2,YMax2,RelationLabel]*\n\nbut in the VRD_sample_submission.csv is like:\n\n    [float,LabelName1,XMin1,XMax1,YMin1,YMax1,LabelName2(or attribute),XMin2,XMax2,YMin2,YMax2,RelationLabel]*\n\nHow should I understand the first float value in each tuple, is the probability of each tuple of relationships?",
    "362979": "The first float is the confidence score -&gt; how much your model is convinced that this particular prediction is correct (take a look at the metrics use in this competition - they make use of this information).",
    "362981": "thanks a lot."
  },
  "source": "meta"
}