{
  "fetched_at": "2026-10-19T11:27:27Z",
  "source_url": "https://www.kaggle.com/competitions/cuhkml-2025-classification/data",
  "pages": [
    {
      "name": "data-description",
      "content": "Dataframe_train and dataframe_test are the main dataset. The difference is that for the test data set, the “action type” (for the classification task) and “expected use time” (for the regression task) variables are masked, and you are asked to predict them.\n\nThe data was already pre-processed. Below is some brief overview of the variables, the rest can be explored on your own.\n\n## Variables\n\n*  ’courier_id’, ’wave_index’, ’tracking_id’, ’date’, ’group’, ’id’ - These are the demographic information of the order and its courier. The column names are somewhat self-explanatory.\n\n* ’courier_wave_start_lng’, ’courier_wave_start_lat’ - These are the starting longitude and latitude of that wave of a certain courier.\n\n* ’level’, ’speed’, ’max_load’ - These are the courier information: The level of the courier, the speed of the courier and the max load of the courier.\n\n* ’weather_grade’ - This is the weather condition of the order.\n\n* ’aoi_id’ - The id of the Area of Interest (i.e. the delivery destination).\n\n* ’shop_id’ - The id of the shop.\n\n* ’source_type’, ’source_tracking_id’, ’source_lng’, ’source_lat’ - The information of the courier’s previous action.\n\n* ’target_lng’, ’target_lat’ - The geographical information of the target.\n\n* ’grid_distance’ - The shortest traversable distance to the target provided by the GPS.\n\n* ’hour’ - The hour in the day.\n\n* ’urgency’ - Identifies how urgent the order is\n\n* ’action_type’ - The outcome variable in the classification task. This is masked in the test set.\n\n* ’expected_use_time’ - The outcome variable in the regression task. This is also masked in the test set\n\nFinally, we want to remind you that NOT ALL VARIABLES ARE NECESSARILY NEEDED. So, you should use your own judgment to decide what to use for your model.\n\n## Files\n\n*   **dataframe_train.csv** - the training set\n*   **dataframe_test.csv** - the test set\n*   **classification_example.csv** - a sample submission file in the correct format\n*   **questions_project.pdf** - Q&A",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "Within the context of Eleme delivery service, at every moment the command platform of \"Smart Logistics\" sends orders from customers to delivery-men for instant delivery. The decision being made by the delivery-men are mainly two-fold: pick-up in store, and delivery to customers. At one specific moment, the delivery-man may receive orders assigned to him, while he has unfinished orders which are previously assigned. He would need to make a decision as to what to do next, based on his order status and geographical location. The goal is to build a model to predict the delivery-man's next move, based on his historical decision and current status.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "The evaluation metric for this competition is Mean F1-Score. The F1 score, commonly used in information retrieval, measures accuracy using the statistics precision p and recall r. Precision is the ratio of true positives (tp) to all predicted positives (tp + fp). Recall is the ratio of true positives to all actual positives (tp + fn). The F1 score is given by [https://en.wikipedia.org/wiki/F-score](https://en.wikipedia.org/wiki/F-score) The 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. Submission Format For every item in the dataset, submission files should contain one column called ID, with a sequence from 0 to 25467. and a column of predicted action_type . There are two types of actions: \"DELIVERY\" and \"PICKUP\". We want you to label \"DELIVERY\" as 1 and \"PICKUP\" as 0. The file should have a header, \"ID\" and \"action_type\" for the two columns, and the file should have 25468 rows in total. The submission should be in CSV format.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "You are allowed to discuss with anyone in this world about this problem set. You are also allowed to read your classmate’s solution. However, what you submit for credit have to be written entirely by yourself.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}