{
  "fetched_at": "2026-10-19T22:23:58Z",
  "source_url": "https://www.kaggle.com/competitions/final-project-stats-415-fall-2023/data",
  "pages": [
    {
      "name": "data-description",
      "content": "*   **X_train.csv** holds features from the training set.\n*  **y_train.csv** holds responses from the training set.\n*   **X_test.csv** holds features from the test set (but the responses are secret!).\n*   **y_sample.csv** holds a sample submission file in the correct format, indicating guesses about each sample in the test set.  When you submit, your file should look like this (but with different values for the \"y\" column)",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "Be nice and try to have fun.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "Welcome to the STATS415 Fa2023 Kaggle competition!\n\nThis competition concerns a (fictional) study, seeking to understand the ingredients that influence student performance on a standardized test of US History.  A large collection of high school students were sampled from 7 school districts.  For each student, the following forms of data were collected.\n\n* `self_eval` -- the student's evaluation of their own mastery of the content.\n* `teacher_eval` -- the student's teacher's evaluation of the student's mastery of the content.\n* `district` -- which school district the student came from.\n* `SRP_1`,`SRP_2`,...`SRP_50` -- a collection of 50 measurements of SRP blood concentration.  SRP is a kind of protein found throughout human tissues.     The student's SRP concentration was measured using a non-invasive optical method (similar to oximetry).  It was measured each day for the fifty days preceding the test, resulting in 50 columns in all.\n* `y` -- the student's performance on the standardized test.\n\nYour task is to understand how the input variables (the self-evaluation, the teacher-evaluation, the district, and the SRP measurements) relate to the response (student performance).  Specifically, you will attempt to learn the conditional expectation of student performance given the input variables.  Your estimated conditional expectation will be evaluated on a secret held-out test-set.\n\nHere's how it works.\n* We give you `X_train.csv` and `y_train.csv`, holding the input features and responses for a large number of students.  Each csv file also includes a `SEQN` column, giving a unique identifier for each student.\n* We give you a `X_test.csv` file holding the input features for a held-out set of students.\n* You must provide us with a `y_test.csv` file, indicating your predictions for performance among the held-out students.  \n\nThe files can be found in the \"Data\" tab above.  You can also find an `y_sample.csv` file indicating the format your answer should take.\n\nYou can submit up to *five times per day*, using the \"Submit Predictions\" button towards the top of this page.  Each time your group submits, you will learn the R-squared of your predictions on the secret data.  For full credit, you must get an R-squared of at least 0.75.  You can get half-credit if you get an R-squared of at least 0.5.  The points on this are a relatively small fraction of the group project score; don't go crazy on this unless you're having fun.  \n\nDon't hesitate to contact GSI Gabriel (gapatron@umich.edu) if you have any questions.\n\nPS. By the way, despite what the progress bar below seems to indicate, the competition did not, in fact, start 2 months ago.  The competition starts at 9am on November 12th.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}