{
  "id": 53567,
  "title": "Data Scientist and a child",
  "url": "/competitions/talkingdata-adtracking-fraud-detection/discussion/53567",
  "author_name": "",
  "post_date": "2018-04-01T17:04:58.992315500Z",
  "votes": 6,
  "comment_count": 6,
  "views": 0,
  "content": "<p>This topic builds upon a post by CPMP.</p>\n\n<p>I feel that if one wants to be a good data scientist he has to behave like a child. \nDuring the growth years, a child tries to touch everything, feel everything and possibly be amazed by those things. A child would implement the ideas without thinking about \"what ifs..\" or \"how\". A child's curiosity is so strong that she/he will think about how even the smallest thing works.</p>\n\n<p>I myself have broken many toys during my childhood days just to see what's inside and how they work. </p>\n\n<p>With this data Scientist field, I am learning by implementing whatever I think would be a good idea. Yes, if you are stuck ask a question about how to move forward, but first, google it. It might not be about sharing your idea but maybe about learning. Here, there are so many kind Kagglers who would help, however, at a job, not many will. </p>\n\n<p>I might be wrong as I am also at learning state. Correct me if I am wrong with this analogy. </p>",
  "messages": [
    {
      "id": "307462",
      "postDate": "04/01/2018 17:04:58",
      "content": "<p>This topic builds upon a post by CPMP.</p>\n\n<p>I feel that if one wants to be a good data scientist he has to behave like a child. \nDuring the growth years, a child tries to touch everything, feel everything and possibly be amazed by those things. A child would implement the ideas without thinking about \"what ifs..\" or \"how\". A child's curiosity is so strong that she/he will think about how even the smallest thing works.</p>\n\n<p>I myself have broken many toys during my childhood days just to see what's inside and how they work. </p>\n\n<p>With this data Scientist field, I am learning by implementing whatever I think would be a good idea. Yes, if you are stuck ask a question about how to move forward, but first, google it. It might not be about sharing your idea but maybe about learning. Here, there are so many kind Kagglers who would help, however, at a job, not many will. </p>\n\n<p>I might be wrong as I am also at learning state. Correct me if I am wrong with this analogy. </p>",
      "rawMarkdown": "This topic builds upon a post by CPMP.\n\nI feel that if one wants to be a good data scientist he has to behave like a child. \nDuring the growth years, a child tries to touch everything, feel everything and possibly be amazed by those things. A child would implement the ideas without thinking about \"what ifs..\" or \"how\". A child's curiosity is so strong that she/he will think about how even the smallest thing works.\n\nI myself have broken many toys during my childhood days just to see what's inside and how they work. \n\nWith this data Scientist field, I am learning by implementing whatever I think would be a good idea. Yes, if you are stuck ask a question about how to move forward, but first, google it. It might not be about sharing your idea but maybe about learning. Here, there are so many kind Kagglers who would help, however, at a job, not many will. \n\n\nI might be wrong as I am also at learning state. Correct me if I am wrong with this analogy.",
      "votes": null
    },
    {
      "id": "307601",
      "postDate": "04/02/2018 05:19:54",
      "content": "<p>Good analogy though .. You should keep trying new things and thats how you learn ! :) </p>",
      "rawMarkdown": "Good analogy though .. You should keep trying new things and thats how you learn ! :)",
      "votes": null
    },
    {
      "id": "307634",
      "postDate": "04/02/2018 06:27:27",
      "content": "<p>I agree with the analogy. Moreover, I believe it is not unique for data scientists though, but for learning in general. Children don't have stereotypes and limiting thoughts in their minds like \"you cannot do this because..\", \"you are too old to learn this\", \"this has to be done in that way\", etc. </p>",
      "rawMarkdown": "I agree with the analogy. Moreover, I believe it is not unique for data scientists though, but for learning in general. Children don't have stereotypes and limiting thoughts in their minds like \"you cannot do this because..\", \"you are too old to learn this\", \"this has to be done in that way\", etc.",
      "votes": null
    },
    {
      "id": "307847",
      "postDate": "04/02/2018 15:29:46",
      "content": "<p>Exactly! There is no upper limit to learning. </p>",
      "rawMarkdown": "Exactly! There is no upper limit to learning.",
      "votes": null
    },
    {
      "id": "307848",
      "postDate": "04/02/2018 15:30:07",
      "content": "<p>yes!</p>",
      "rawMarkdown": "yes!",
      "votes": null
    },
    {
      "id": "308084",
      "postDate": "04/02/2018 23:23:14",
      "content": "<p>I like the analogy, but I feel it is limited to learning / side projects. As a full-time data scientist, I do think I have a lot more room to explore, compared with a software engineer. But I wouldn't say I can \"implement whatever I think would be a good idea\". Projects have scope, deadlines and expectations. It's unrealistic to go down every single rabbit hole. </p>\n\n<p>The way I'd like to think about data science is like a constrained optimization problem. Instead of an EDA that leaves no rock unturned or a \"bigger is better / throw all shit at the wall to see what sticks\" model ensemble, I see it as \"EDA under 4 hours\" or \"best models under 5 attempts\".</p>",
      "rawMarkdown": "I like the analogy, but I feel it is limited to learning / side projects. As a full-time data scientist, I do think I have a lot more room to explore, compared with a software engineer. But I wouldn't say I can \"implement whatever I think would be a good idea\". Projects have scope, deadlines and expectations. It's unrealistic to go down every single rabbit hole. \n\nThe way I'd like to think about data science is like a constrained optimization problem. Instead of an EDA that leaves no rock unturned or a \"bigger is better / throw all shit at the wall to see what sticks\" model ensemble, I see it as \"EDA under 4 hours\" or \"best models under 5 attempts\".",
      "votes": null
    },
    {
      "id": "309706",
      "postDate": "04/05/2018 20:53:00",
      "content": "<p>Thank you for bringing in a different POV. I agree with that. However, what if someone isn't that experienced and an aspiring Data Scientist?</p>",
      "rawMarkdown": "Thank you for bringing in a different POV. I agree with that. However, what if someone isn't that experienced and an aspiring Data Scientist?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 307601,
      "author_name": "prashanththangavel",
      "author_url": "",
      "post_date": "04/02/2018 05:19:54",
      "content": "<p>Good analogy though .. You should keep trying new things and thats how you learn ! :) </p>",
      "votes": null,
      "replies": [
        {
          "id": 307848,
          "author_name": "aroragaurav",
          "author_url": "",
          "post_date": "04/02/2018 15:30:07",
          "content": "<p>yes!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 307634,
      "author_name": "asparuhhristov",
      "author_url": "",
      "post_date": "04/02/2018 06:27:27",
      "content": "<p>I agree with the analogy. Moreover, I believe it is not unique for data scientists though, but for learning in general. Children don't have stereotypes and limiting thoughts in their minds like \"you cannot do this because..\", \"you are too old to learn this\", \"this has to be done in that way\", etc. </p>",
      "votes": null,
      "replies": [
        {
          "id": 307847,
          "author_name": "aroragaurav",
          "author_url": "",
          "post_date": "04/02/2018 15:29:46",
          "content": "<p>Exactly! There is no upper limit to learning. </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 308084,
      "author_name": "sijunhe9248",
      "author_url": "",
      "post_date": "04/02/2018 23:23:14",
      "content": "<p>I like the analogy, but I feel it is limited to learning / side projects. As a full-time data scientist, I do think I have a lot more room to explore, compared with a software engineer. But I wouldn't say I can \"implement whatever I think would be a good idea\". Projects have scope, deadlines and expectations. It's unrealistic to go down every single rabbit hole. </p>\n\n<p>The way I'd like to think about data science is like a constrained optimization problem. Instead of an EDA that leaves no rock unturned or a \"bigger is better / throw all shit at the wall to see what sticks\" model ensemble, I see it as \"EDA under 4 hours\" or \"best models under 5 attempts\".</p>",
      "votes": null,
      "replies": [
        {
          "id": 309706,
          "author_name": "aroragaurav",
          "author_url": "",
          "post_date": "04/05/2018 20:53:00",
          "content": "<p>Thank you for bringing in a different POV. I agree with that. However, what if someone isn't that experienced and an aspiring Data Scientist?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "307462": "This topic builds upon a post by CPMP.\n\nI feel that if one wants to be a good data scientist he has to behave like a child. \nDuring the growth years, a child tries to touch everything, feel everything and possibly be amazed by those things. A child would implement the ideas without thinking about \"what ifs..\" or \"how\". A child's curiosity is so strong that she/he will think about how even the smallest thing works.\n\nI myself have broken many toys during my childhood days just to see what's inside and how they work. \n\nWith this data Scientist field, I am learning by implementing whatever I think would be a good idea. Yes, if you are stuck ask a question about how to move forward, but first, google it. It might not be about sharing your idea but maybe about learning. Here, there are so many kind Kagglers who would help, however, at a job, not many will. \n\n\nI might be wrong as I am also at learning state. Correct me if I am wrong with this analogy.",
    "307601": "Good analogy though .. You should keep trying new things and thats how you learn ! :)",
    "307634": "I agree with the analogy. Moreover, I believe it is not unique for data scientists though, but for learning in general. Children don't have stereotypes and limiting thoughts in their minds like \"you cannot do this because..\", \"you are too old to learn this\", \"this has to be done in that way\", etc.",
    "307847": "Exactly! There is no upper limit to learning.",
    "307848": "yes!",
    "308084": "I like the analogy, but I feel it is limited to learning / side projects. As a full-time data scientist, I do think I have a lot more room to explore, compared with a software engineer. But I wouldn't say I can \"implement whatever I think would be a good idea\". Projects have scope, deadlines and expectations. It's unrealistic to go down every single rabbit hole. \n\nThe way I'd like to think about data science is like a constrained optimization problem. Instead of an EDA that leaves no rock unturned or a \"bigger is better / throw all shit at the wall to see what sticks\" model ensemble, I see it as \"EDA under 4 hours\" or \"best models under 5 attempts\".",
    "309706": "Thank you for bringing in a different POV. I agree with that. However, what if someone isn't that experienced and an aspiring Data Scientist?"
  },
  "source": "meta"
}