{
  "id": 571438,
  "title": "Do we have only one test image!?? ",
  "url": "/competitions/el-hackathon-2025/discussion/571438",
  "author_name": "",
  "post_date": "2025-04-03T08:22:02.516975900Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Evaluating a method using just one test image is nothing more than a gamble!</p>",
  "messages": [
    {
      "id": "3169211",
      "postDate": "04/03/2025 08:22:02",
      "content": "<p>Evaluating a method using just one test image is nothing more than a gamble!</p>",
      "rawMarkdown": "Evaluating a method using just one test image is nothing more than a gamble!",
      "votes": null
    },
    {
      "id": "3169223",
      "postDate": "04/03/2025 08:38:51",
      "content": "<p>For that matter, we also have only 6 images for the training set, which is not a lot.</p>\n<p>However, we are predicting multiple spots on the slides, and S_7 seems to have some heterogeneous areas, which may have a different composition than the rest.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F650%2Ff72d5ab338609819f11549865d4c6159%2F__results___16_1%20(1).png?generation=1743615424148607&amp;alt=media\" alt=\"\"><br>\n(darker dots are the LB Test set, ligher ones are the private LB)</p>",
      "rawMarkdown": "For that matter, we also have only 6 images for the training set, which is not a lot.\n\nHowever, we are predicting multiple spots on the slides, and S_7 seems to have some heterogeneous areas, which may have a different composition than the rest.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F650%2Ff72d5ab338609819f11549865d4c6159%2F__results___16_1%20(1).png?generation=1743615424148607&alt=media)\n(darker dots are the LB Test set, ligher ones are the private LB)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3169223,
      "author_name": "dalloliogm",
      "author_url": "",
      "post_date": "04/03/2025 08:38:51",
      "content": "<p>For that matter, we also have only 6 images for the training set, which is not a lot.</p>\n<p>However, we are predicting multiple spots on the slides, and S_7 seems to have some heterogeneous areas, which may have a different composition than the rest.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F650%2Ff72d5ab338609819f11549865d4c6159%2F__results___16_1%20(1).png?generation=1743615424148607&amp;alt=media\" alt=\"\"><br>\n(darker dots are the LB Test set, ligher ones are the private LB)</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3169211": "Evaluating a method using just one test image is nothing more than a gamble!",
    "3169223": "For that matter, we also have only 6 images for the training set, which is not a lot.\n\nHowever, we are predicting multiple spots on the slides, and S_7 seems to have some heterogeneous areas, which may have a different composition than the rest.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F650%2Ff72d5ab338609819f11549865d4c6159%2F__results___16_1%20(1).png?generation=1743615424148607&alt=media)\n(darker dots are the LB Test set, ligher ones are the private LB)"
  },
  "source": "meta"
}