{
  "id": 407730,
  "title": "biological background",
  "url": "/competitions/cafa-5-protein-function-prediction/discussion/407730",
  "author_name": "",
  "post_date": "2023-05-07T22:12:58.461348900Z",
  "votes": 5,
  "comment_count": 1,
  "views": 0,
  "content": "<p>this competition need biological background or a solid understanding ?</p>",
  "messages": [
    {
      "id": "2249581",
      "postDate": "05/07/2023 22:12:58",
      "content": "<p>this competition need biological background or a solid understanding ?</p>",
      "rawMarkdown": "this competition need biological background or a solid understanding ?",
      "votes": null
    },
    {
      "id": "2264646",
      "postDate": "05/18/2023 15:44:30",
      "content": "<p>I think that if you have some background it is definitely useful. However, there are many things that can get you 'up-to-speed' quickly to a level that can be beneficial for your submissions: </p>\n<ul>\n<li>Read papers on the matter and how researchers implemented their models to solve this specific one</li>\n<li>Find discussions here where useful resources are linked and take some time to read them</li>\n<li>As was said in one of the discussions, language models embeddings can take you far even with zero knowledge on the subject</li>\n</ul>\n<p>The difficult part here is dealing with validation, class inbalance and sparsity, label and feature correlations and finally representing the GO graph in your final predictions + making this fit into the notebook kernel 💥 </p>\n<p>Starting though should not be too difficult!</p>",
      "rawMarkdown": "I think that if you have some background it is definitely useful. However, there are many things that can get you 'up-to-speed' quickly to a level that can be beneficial for your submissions: \n- Read papers on the matter and how researchers implemented their models to solve this specific one\n- Find discussions here where useful resources are linked and take some time to read them\n- As was said in one of the discussions, language models embeddings can take you far even with zero knowledge on the subject\n\nThe difficult part here is dealing with validation, class inbalance and sparsity, label and feature correlations and finally representing the GO graph in your final predictions + making this fit into the notebook kernel 💥 \n\nStarting though should not be too difficult!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2264646,
      "author_name": "wadgek",
      "author_url": "",
      "post_date": "05/18/2023 15:44:30",
      "content": "<p>I think that if you have some background it is definitely useful. However, there are many things that can get you 'up-to-speed' quickly to a level that can be beneficial for your submissions: </p>\n<ul>\n<li>Read papers on the matter and how researchers implemented their models to solve this specific one</li>\n<li>Find discussions here where useful resources are linked and take some time to read them</li>\n<li>As was said in one of the discussions, language models embeddings can take you far even with zero knowledge on the subject</li>\n</ul>\n<p>The difficult part here is dealing with validation, class inbalance and sparsity, label and feature correlations and finally representing the GO graph in your final predictions + making this fit into the notebook kernel 💥 </p>\n<p>Starting though should not be too difficult!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2249581": "this competition need biological background or a solid understanding ?",
    "2264646": "I think that if you have some background it is definitely useful. However, there are many things that can get you 'up-to-speed' quickly to a level that can be beneficial for your submissions: \n- Read papers on the matter and how researchers implemented their models to solve this specific one\n- Find discussions here where useful resources are linked and take some time to read them\n- As was said in one of the discussions, language models embeddings can take you far even with zero knowledge on the subject\n\nThe difficult part here is dealing with validation, class inbalance and sparsity, label and feature correlations and finally representing the GO graph in your final predictions + making this fit into the notebook kernel 💥 \n\nStarting though should not be too difficult!"
  },
  "source": "meta"
}