{
  "fetched_at": "2026-10-20T07:05:10Z",
  "source_url": "https://www.kaggle.com/competitions/DSAA5013Spring/data",
  "pages": [
    {
      "name": "data-description",
      "content": "# Dataset Description\n\n## File descriptions\n\n* **train.csv** - the training set (2,000 samples)\n* **test.csv** - the test set (1,000 samples)\n* **samplesubmission.csv** - a sample submission file in the correct format\n\n## Columns of training data\n\n* **DEFAULT** - 1 if defaulted 0 if not\n* **AGE - Age** in years plus twelfths of a year\n* **ACADMOS** - months living at current address\n* **ADEPCNT** - 1 + number of dependents\n* **MAJORDRG** - Number of major derogatory reports\n* **MINORDRG** - Number of minor derogatory reports\n* **OWNRENT** - 1 if owns their home, 0 if rent\n* **NCOME** - Monthly income (divided by 10,000)\n* **SELFEMPL** - 1 if self employed, 0 if not\n* **INCPER** - Income divided by number of dependents\n* **EXP_INC** - Ratio of monthly credit card expenditure to yearly income\n* **SPENDING** - Average monthly credit card expenditure\n* **LOGSPEND** - Log of spending",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "# Project description\n\nCredit default is critical to the continuous operation of commercial banks. A retail bank is expecting a credit default model to identify the consumers who are likely to default on their credit card payments over the next 12 months. This model can be used for future actions taken to manage the bank's risk and reduce future losses.\n\nSome sample data extracted from their systems are available for you to build a credit default model and predict likely default consumers. Based on the bank’s experience, the number of derogatory reports is a strong indicator of default. This is all the information you are able to get from the bank at the moment.\n\nYou can choose any model you want to do this task, and you are suggested to learn some basic knowledge of banking and get familiar with its terminology. Data preprocessing is needed, and you may want to do some feature engineering (like normalization or discretization). All kinds of information related to the bank model are welcome to refer to.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "# Rules\n\n - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "# Evaluation Metric \n\nThe evaluation metric for this Kaggle competition is [Mean F1-Score](https://en.wikipedia.org/wiki/F-score). The F1 score, commonly used in information retrieval, measures accuracy using the precision of the statistics \\( \\text{p} \\) and recall \\( \\text{r} \\). Precision is the ratio of true positives \\( \\text{tp} \\) to all predicted positives \\( \\text{tp+fp} \\). The recall is the ratio of true positives \\( \\text{tp} \\) to all actual positives \\( \\text{tp+fp} \\). The F1 score is given by:\n\n\\[ \\text{F1} = 2\\frac{\\text{p} \\cdot \\text{r}}{\\text{p}+\\text{r}}\\ \\ \\mathrm{where}\\ \\ \\text{p} = \\frac{\\text{tp}}{\\text{tp}+\\text{fp}},\\ \\ \\text{r} = \\frac{\\text{tp}}{\\text{tp}+\\text{fn}} \\]\n\nThe F1 metric weights recall and precision equally, and a good retrieval algorithm will maximize both precision and recall simultaneously. Thus, moderately good performance on both will be favored over extremely good performance on one and poor performance on the other. In your report, you can use various metrics to demonstrate your prediction results and you need to briefly describe the metrics you choose. Matrices and diagrams are also welcomed with your analysis.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}