{
  "id": 85193,
  "title": "My Best Submissions (Public 0.365 → Private 0.694)",
  "url": "/competitions/vsb-power-line-fault-detection/writeups/uratatsu-my-best-submissions-public-0-365-private-",
  "author_name": "",
  "post_date": "2019-03-22T08:00:21.571532200Z",
  "votes": 3,
  "comment_count": 1,
  "views": 0,
  "content": "<p>My best private score is 0.694(gold rank), but public score is 0.365 and 0.511.\nWhat on earth is that? </p>\n\n<ol>\n<li>LGBM with basic statistics(mean,minmax,std, etc.) of raw signals +  basic statistics of FFT denoised signals.</li>\n</ol>\n\n<p>Public    Private\n0.365 → 0.694</p>\n\n<p>Local CV\n0.63991</p>\n\n<ol>\n<li>LGBM with basic statistics(mean,minmax,std, etc.) of raw signals and features of Host’s papers (Mean width of peaks ,Max width of peaks  etc.) of FFT denoised signals.</li>\n</ol>\n\n<p>Public    Private\n0.511 → 0.694</p>\n\n<p>Local CV\n0.67035</p>\n\n<p>I can’t select these submissions. It’s so difficult.</p>",
  "messages": [
    {
      "id": "496439",
      "postDate": "03/22/2019 08:00:21",
      "content": "<p>My best private score is 0.694(gold rank), but public score is 0.365 and 0.511.\nWhat on earth is that? </p>\n\n<ol>\n<li>LGBM with basic statistics(mean,minmax,std, etc.) of raw signals +  basic statistics of FFT denoised signals.</li>\n</ol>\n\n<p>Public    Private\n0.365 → 0.694</p>\n\n<p>Local CV\n0.63991</p>\n\n<ol>\n<li>LGBM with basic statistics(mean,minmax,std, etc.) of raw signals and features of Host’s papers (Mean width of peaks ,Max width of peaks  etc.) of FFT denoised signals.</li>\n</ol>\n\n<p>Public    Private\n0.511 → 0.694</p>\n\n<p>Local CV\n0.67035</p>\n\n<p>I can’t select these submissions. It’s so difficult.</p>",
      "rawMarkdown": "My best private score is 0.694(gold rank), but public score is 0.365 and 0.511.\nWhat on earth is that? \n\n1. LGBM with basic statistics(mean,minmax,std, etc.) of raw signals +  basic statistics of FFT denoised signals.\n \nPublic    Private\n0.365 → 0.694\n\nLocal CV\n0.63991\n\n2. LGBM with basic statistics(mean,minmax,std, etc.) of raw signals and features of Host’s papers (Mean width of peaks ,Max width of peaks  etc.) of FFT denoised signals.\n\nPublic    Private\n0.511 → 0.694\n\nLocal CV\n0.67035\n\nI can’t select these submissions. It’s so difficult.",
      "votes": null
    },
    {
      "id": "496684",
      "postDate": "03/22/2019 13:17:11",
      "content": "<p>Encounter the same problem. Extreme low public score makes me feel it's overfitted, then filtered it out...    : )</p>",
      "rawMarkdown": "Encounter the same problem. Extreme low public score makes me feel it's overfitted, then filtered it out...    : )",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 496684,
      "author_name": "gitshe11",
      "author_url": "",
      "post_date": "03/22/2019 13:17:11",
      "content": "<p>Encounter the same problem. Extreme low public score makes me feel it's overfitted, then filtered it out...    : )</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "496439": "My best private score is 0.694(gold rank), but public score is 0.365 and 0.511.\nWhat on earth is that? \n\n1. LGBM with basic statistics(mean,minmax,std, etc.) of raw signals +  basic statistics of FFT denoised signals.\n \nPublic    Private\n0.365 → 0.694\n\nLocal CV\n0.63991\n\n2. LGBM with basic statistics(mean,minmax,std, etc.) of raw signals and features of Host’s papers (Mean width of peaks ,Max width of peaks  etc.) of FFT denoised signals.\n\nPublic    Private\n0.511 → 0.694\n\nLocal CV\n0.67035\n\nI can’t select these submissions. It’s so difficult.",
    "496684": "Encounter the same problem. Extreme low public score makes me feel it's overfitted, then filtered it out...    : )"
  },
  "source": "meta"
}