{
  "fetched_at": "2026-10-22T10:27:23Z",
  "source_url": "https://www.kaggle.com/competitions/adml-final/data",
  "pages": [
    {
      "name": "data-description",
      "content": "Data is included in the file \n\n**Train.csv** - The training set\n\n**Test.csv** - The test set\n\nEach row represents a customer, each column contains following customer’s attributes \n\n- 1 - `age`  \n\n- 2 - `job`: type of job \n\n- 3 - `marital`: marital status \n\n- 4 - `education` \n\n- 5 - `default`: has credit in default? \n\n- 6 - `housing`: has housing loan? \n\n- 7 - `loan`: has personal loan? \n\n- 8 - `balance`: Balance of the individual.\n\n- 9 - `contact`: contact communication type\n \n- 10 - `month`: last contact month of year\n \n- 11 - `day`: last contact day of the month\n \n- 12 - `campaign`: number of contacts performed during this campaign and for this customer (numeric, includes last contact)\n\n- 13 - `pdays`: number of days that passed by after the customer was last contacted from a previous campaign (numeric; -1 means customer was not previously contacted)\n\n- 14 - `previous`: number of contacts performed before this campaign and for this customer (numeric)\n\n- 15 - `poutcome`: outcome of the previous marketing campaign (categorical: `'failure'`, `'nonexistent'`,`'success'`)\n\nOutput variable (desired target):\n\n- 16 - `y` - has the customer purchased the product? (binary: `'yes'`,`'no'`)",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Problem Statement",
      "content": "You work at a multinational bank that is aiming to increase it's market share. Recently, the bank have been promoting one of it’s core products. A key issue is optimizing targeting for marketing is, under a growing pressure to increase profits and reduce costs.\n\nAs a data scientist, you are working on improving for the next marketing campaign with a greater effectiveness and ROI. In order to achieve this, we have to analyze the past consumer data, identify the patterns, and predict the consumer’s propensity to buy the promoted product. The model result is a valuable tool to help the marketing team to use their budget wisely to target potential buyers. \n \nWith the given training data, you need to train and fine-tune a series of models, and then apply the model to test data and predict the outcome. The ROC AUC will be used to evaluate the model performance\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": " - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Questions",
      "content": "\n## Questions:\n\n### 1.\t(10 points) Import customer data. Do some basic exploratory data analysis to describe the data numerically or graphically. \n### 2.\t(20 points) Based on training data, build and fine-tune a logistic regression model to predict if a customer will purchase or not. Please evaluate the model performance using cross-validation. \n### 3.\t(20 points) Based on training data, build and fine-tune a catboost or xgboost model to predict if a customer will purchase or not. Please evaluate the model performance using cross-validation. Hint: Catboost has an internal encoding algorithm to handle categorical variables.  \n### 4.\t(10 points) Based on training data, build and fine-tune another non-tree based model, e.g., MLP, SVM etc. Please evaluate the model performance using cross-validation\n### 5.\t(20 points) Propose and build a stacking model based on previous models you have built. \n### 6.\t(10 points) Based on previous models you have built, apply the model to test data and predict the outcome. \n### 7.\t(10 points) If the marketing department want to use the prediction in question 6 directly to select the potential customers for the campaign, what potential problems do you see from this approach? What methods do you suggest to select the potential customers?  ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "See data description under the \"Data\" Tag.\nPlease rename your submission file with:  netID + \"_final.csv\"",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}