📦 Dataset: Bank Marketing (bank-additional-full.csv)

The dataset is from a Portuguese banking institution, collected from 2008–2013.

It contains information from phone-based marketing campaigns aimed at selling term deposits.

Includes categorical and numerical attributes related to customer profiles, campaign success, and economic context.

More details: 📄 UCI Bank Marketing Dataset

• Source: UCI Machine Learning Repository
• Goal: Predict whether a client will subscribe to a term deposit
• Target Variable: y (values: "yes" or "no")
• Records: 41,188
• Features: 20+ (including age, job, education, duration, poutcome, emp.var.rate)

---

🔍 Preprocessing

• Converted all character columns to factors
• Split data:
  • 80% training
  • 20% testing
• Reproducible with set.seed(123)

---

🌲 Model: Random Forest Classifier

• Trained on all features to predict y
• Parameters:
  • ntree = 3

---

🖥️ Shiny Dashboard Features

📈 Sidebar

• Accuracy
• Precision
• Recall
• F1 Score

🔍 Tabs

• Confusion Matrix: Heatmap using ggplot2
• Feature Importance: Table view of top variables
• Predictions: First 6 predictions with total count
• ROC Curve: Plots True Positive Rate vs. False Positive Rate

---

📁 Project Structure

/RANDOM-FOREST
├── data/
│   └── bank-additional-full.csv
├── random-forest-model.r   # model training, predictions, metrics
├── app.r                   # Shiny dashboard consuming results
├── renv/                   # R environment tracking

---

📌 Notes

• All data and model objects are stored in rf_dashboard (a shared environment).
• random-forest-model.r is sourced inside app.r to keep logic modular.
• No model is saved — everything runs in memory.

---