{
  "id": 231007,
  "title": "Anyone try interpolation for waypoint in wifi dataset?",
  "url": "/competitions/indoor-location-navigation/discussion/231007",
  "author_name": "",
  "post_date": "2021-04-06T13:29:36.085858800Z",
  "votes": 7,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hi, everyone!<br>\nI shared <a href=\"https://www.kaggle.com/kuto0633/linear-interpolation-for-waypoint-in-wifi-dataset\" target=\"_blank\">the notebook</a> which explain how to get a more correct waypoint corresponding to wifi timestamp by linear interpolation.</p>\n<p>The visualization results show that linear interpolation seems to work well, but the CV and LB score didn't improve as much as I expected.(about 0.1) </p>\n<p>Has anyone else tried like this or other interpolation method? And what was the result?</p>",
  "messages": [
    {
      "id": "1264869",
      "postDate": "04/06/2021 13:29:36",
      "content": "<p>Hi, everyone!<br>\nI shared <a href=\"https://www.kaggle.com/kuto0633/linear-interpolation-for-waypoint-in-wifi-dataset\" target=\"_blank\">the notebook</a> which explain how to get a more correct waypoint corresponding to wifi timestamp by linear interpolation.</p>\n<p>The visualization results show that linear interpolation seems to work well, but the CV and LB score didn't improve as much as I expected.(about 0.1) </p>\n<p>Has anyone else tried like this or other interpolation method? And what was the result?</p>",
      "rawMarkdown": "Hi, everyone!\nI shared [the notebook](https://www.kaggle.com/kuto0633/linear-interpolation-for-waypoint-in-wifi-dataset) which explain how to get a more correct waypoint corresponding to wifi timestamp by linear interpolation.\n  \nThe visualization results show that linear interpolation seems to work well, but the CV and LB score didn't improve as much as I expected.(about 0.1) \n\nHas anyone else tried like this or other interpolation method? And what was the result?",
      "votes": null
    },
    {
      "id": "1265863",
      "postDate": "04/07/2021 09:07:14",
      "content": "<p>Do you mean increasing the number of the training samples by interpolating waypoints? Sounds good, I will try this :)</p>",
      "rawMarkdown": "Do you mean increasing the number of the training samples by interpolating waypoints? Sounds good, I will try this :)",
      "votes": null
    },
    {
      "id": "1265905",
      "postDate": "04/07/2021 09:57:52",
      "content": "<p><a href=\"https://www.kaggle.com/mamasinkgs\" target=\"_blank\">@mamasinkgs</a> That's right.  I'm looking forward to your result :)</p>",
      "rawMarkdown": "mamasinkgs That's right.  I'm looking forward to your result :)",
      "votes": null
    },
    {
      "id": "1266271",
      "postDate": "04/07/2021 16:01:21",
      "content": "<p>I think the compute_step_positions function provided by the host interpolates waypoints using sensor data. <a href=\"https://github.com/location-competition/indoor-location-competition-20/blob/master/compute_f.py\" target=\"_blank\">https://github.com/location-competition/indoor-location-competition-20/blob/master/compute_f.py</a>.</p>",
      "rawMarkdown": "I think the compute_step_positions function provided by the host interpolates waypoints using sensor data. https://github.com/location-competition/indoor-location-competition-20/blob/master/compute_f.py.",
      "votes": null
    },
    {
      "id": "1266665",
      "postDate": "04/08/2021 01:21:47",
      "content": "<p><a href=\"https://www.kaggle.com/nooblife\" target=\"_blank\">@nooblife</a> Thanks for teaching me. I’ll check it.</p>",
      "rawMarkdown": "nooblife Thanks for teaching me. I’ll check it.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1265863,
      "author_name": "mamasinkgs",
      "author_url": "",
      "post_date": "04/07/2021 09:07:14",
      "content": "<p>Do you mean increasing the number of the training samples by interpolating waypoints? Sounds good, I will try this :)</p>",
      "votes": null,
      "replies": [
        {
          "id": 1265905,
          "author_name": "kuto0633",
          "author_url": "",
          "post_date": "04/07/2021 09:57:52",
          "content": "<p><a href=\"https://www.kaggle.com/mamasinkgs\" target=\"_blank\">@mamasinkgs</a> That's right.  I'm looking forward to your result :)</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1266271,
      "author_name": "nooblife",
      "author_url": "",
      "post_date": "04/07/2021 16:01:21",
      "content": "<p>I think the compute_step_positions function provided by the host interpolates waypoints using sensor data. <a href=\"https://github.com/location-competition/indoor-location-competition-20/blob/master/compute_f.py\" target=\"_blank\">https://github.com/location-competition/indoor-location-competition-20/blob/master/compute_f.py</a>.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1266665,
          "author_name": "kuto0633",
          "author_url": "",
          "post_date": "04/08/2021 01:21:47",
          "content": "<p><a href=\"https://www.kaggle.com/nooblife\" target=\"_blank\">@nooblife</a> Thanks for teaching me. I’ll check it.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1264869": "Hi, everyone!\nI shared [the notebook](https://www.kaggle.com/kuto0633/linear-interpolation-for-waypoint-in-wifi-dataset) which explain how to get a more correct waypoint corresponding to wifi timestamp by linear interpolation.\n  \nThe visualization results show that linear interpolation seems to work well, but the CV and LB score didn't improve as much as I expected.(about 0.1) \n\nHas anyone else tried like this or other interpolation method? And what was the result?",
    "1265863": "Do you mean increasing the number of the training samples by interpolating waypoints? Sounds good, I will try this :)",
    "1265905": "mamasinkgs That's right.  I'm looking forward to your result :)",
    "1266271": "I think the compute_step_positions function provided by the host interpolates waypoints using sensor data. https://github.com/location-competition/indoor-location-competition-20/blob/master/compute_f.py.",
    "1266665": "nooblife Thanks for teaching me. I’ll check it."
  },
  "source": "meta"
}