📌 C4.5 Decision Tree – Fast & Scalable Decision Tree in Go

A high-performance and scalable C4.5 decision tree classifier implemented in Go. This tool enables users to efficiently train a decision tree model on structured data and make accurate predictions using the trained model.

🚀 Optimized for speed, parallel execution, and large datasets.

📑 Table of Contents

• 🚀 Features
• ⚙️ How It Works
• 📂 Project Structure
• 📥 Installation
• 🔧 Usage
  • Building the Project
  • Training a Decision Tree
  • Making Predictions
• 📜 License
• 🙌 Contributors
• 🤝 Contributing

---

🚀 Features

✔ CSV Data Processing – Reads CSV files, extracts features, and identifies the target labels.
✔ Parallel Processing – Uses Go goroutines to speed up data handling and decision tree building.
✔ C4.5 Algorithm – Implements the C4.5 decision tree with entropy-based splitting and pruning.
✔ Feature Selection – Selects the best feature at each node to maximize information gain.
✔ Handles Missing Values – Uses smart imputation techniques to handle missing data.
✔ Fast Predictions – Efficiently classifies new data points using the trained decision tree model.
✔ Serialization – Saves trained models as JSON files for later use in predictions.
✔ Command-Line Interface – Simple CLI for training and predicting with decision trees.

---

⚙️ How It Works

1️⃣ Data Processing: Parses CSV files and detects headers.
2️⃣ Feature Selection: Uses entropy and information gain to find the best splits.
3️⃣ Tree Building: Recursively builds the decision tree, using pruning for efficiency.
4️⃣ Model Storage: Saves the trained decision tree in a serializable JSON format.
5️⃣ Predictions: Uses the trained tree to classify new input data.

---

📂 Project Structure

|─ cmd/                # CLI commands and argument parsing  
│   ├── root.go        # CLI entry point for commands  
│  
├── internal/model/    # Core logic for decision tree training and predictions  
│   ├── cache/        # Caches computed values for performance optimization  
│   ├── counter/      # Computes class distributions (e.g., mode in a class)  
│   ├── entropy/      # Calculates data uncertainty (entropy calculation)  
│   ├── model/        # Trains the decision tree based on input data  
│   ├── node/         # Defines tree node structure and utility functions  
│   ├── parser/       # Parses CSV files and converts data into structured format  
│   ├── predict/      # Uses the trained model to make predictions  
│   ├── split/        # Finds the best feature split for information gain  
│   ├── types/        # Defines tree structure and related data types  
│   ├── utils/        # Utility functions for data preprocessing  
│  
├── decision_model/    # Stores serialized trained decision tree models  
├── go.mod             # Go module dependencies  
├── go.sum             # Go dependency checksums  
├── LICENSE            # License information  
├── main.go            # Application entry point  

---

📥 Installation

1️⃣ Install Go (if not installed)

Ensure you have Go installed. 🔗 Download Go

2️⃣ Clone the Repository

git clone https://learn.zone01kisumu.ke/git/tesiaka/c4.5-decision-tree.git
cd c4.5-decision-tree

3️⃣ Install Dependencies

go mod tidy

---

🔧 Usage

Building the Project

go build -o dt

This creates an executable dt for running commands.

---

Training a Decision Tree

Flag	Description
-c	Train a decision tree (train)
-i	Input CSV file path containing the training dataset
-t	Name of the column in the dataset containing the target labels
-o	Output file to save the trained decision tree (JSON format)

Example (training):

./dt -c train -i dataset.csv -t target_column -o model.dt

---

Making Predictions

Flag	Description
-c	Predict command (predict)
-i	Input CSV file containing test data
-m	Path to the trained decision tree model file
-o	Path to save predictions as a CSV file

Example (prediction):

./dt -c predict -i test_data.csv -m model.dt -o predictions.csv

---

📜 License

This project is licensed under the MIT License.
🔗 MIT License

---

🙌 Contributors

1. John Paul
2. Antony Odour
3. Teddy Siaka
4. David Jesse
5. Amos Joel

---

🤝 Contributing

🚀 This project is open for contributions!

🔹 How to contribute:

1. Fork the repository.
2. Create a new branch for your feature.
3. Commit your changes.
4. Push the branch to your fork.
5. Open a pull request.
6. Submit your pull request.
7. Review and merge.
8. Update the documentation.
9. Update the changelog.

💡 Let's build a faster and more efficient Decision Tree model together!

---

Developed with ❤️ in Go. 🚀