{
  "fetched_at": "2026-10-19T11:24:52Z",
  "source_url": "https://www.kaggle.com/competitions/cs420-2019-ctr-prediction/data",
  "pages": [
    {
      "name": "rules",
      "content": "- You are allowed to use machine learning libraries in this course work. \n- Nothing but the given training data should be used to train your model.\n- Do not copy the code from others or Internet sources. You should also submit the code for your model when the competition ends.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "data-description",
      "content": "# File Descriptions\n* train.data - the training set \n* test.data - the test set\n* sample_submission.csv - a sample submission file in the correct format\n\n# Data Fields\nPlease refer to the introduction slides of this course work.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "[Click-through rate (CTR)](https://en.wikipedia.org/wiki/Click-through_rate) is the ratio of users who click on a specific link, or the probability for a specific user to click on a specific advertisement. It is a significant metric for online advertising. In this course work, your task is to develop a model to predict CTRs for the given cases in online advertising.  Modern search engines and recommender systems perform CTR predictions at a high frequency to decide which ads (or items) to display when a user requires a webpage, in which way the platforms try to maximize their revenue.\n\nYou may refer to the introduction slides of this course work for more details. Questions are welcome on the [Discussion](https://www.kaggle.com/c/cs420-2019-ctr-prediction/discussion) page. ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "The evaluation metric used in this competition is AUC (Area Under Receiver Operating Characteristic Curve). You may refer to [this tutorial](https://developers.google.com/machine-learning/crash-course/classification/roc-and-auc) if you are not familiar with ROC and AUC.\n\nYour final score on the private leaderboard is used to calculate your grade on this course work. \n\n# Submission Format\nThe submissions should contain the predicted CTR for each test entry in the following format:\n\n    Id,Predicted\n    0,0.7541380745472263\n    1,0.7694497722718407\n    2,0.17182101447890197\n    3,0.5118826329016144\n    4,0.44676858073715475\n    ...\n\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}