🧬 Genetic Programming for Classification

This repository contains a Java implementation of a Genetic Programming (GP) algorithm for solving binary classification problems. The GP algorithm evolves a population of tree-based models to predict the target variable in a dataset.

---

🌟 Features

• 🌳 Tree-based Representation: Each individual in the population is represented as a binary tree, where leaf nodes are constants and internal nodes are arithmetic operations.
• ⚡ Parallel Evolution: The evolution process is parallelized using Java's ExecutorService for improved performance.
• 🎛️ Customizable Parameters: Adjust key parameters such as population size, mutation rate, and crossover rate.
• 📊 Metrics Calculation: The program calculates accuracy, specificity, sensitivity, and F-measure for the test set.

---

🛠️ Requirements

To run this project, you need:

• Java Development Kit (JDK) 17 or higher
• Visual Studio Code (or any Java IDE)
• A dataset in CSV format (see Dataset for details)

---

📂 Dataset

The program expects a dataset in CSV format with the following characteristics:

• The last column should contain the target variable (binary: 0 or 1).
• All other columns should contain numeric features.
• The first row can be a header row (e.g., column names), but it will be skipped during data loading.

Example Dataset Format

feature1,feature2,feature3,target
0.1,0.5,0.7,1
0.2,0.3,0.8,0
0.4,0.6,0.9,1

---

🗂️ Code Structure

The project consists of the following Java files:

File	Description
GeneticProgramming.java	Contains the main GP algorithm, including population initialization, evolution, and selection.
GPNode.java	Represents a node in the GP tree. Includes methods for tree generation, evaluation, mutation, and crossover.
Main.java	The entry point of the program. Loads the dataset, initializes the GP algorithm, and prints the results.
README.md	This file, providing an overview of the project and instructions for running the code.

---

🚀 How to Run

Step 1: Clone the Repository

git clone https://github.com/your-username/genetic-programming-classification.git
cd genetic-programming-classification

Step 2: Open the Project in VS Code

1. Open Visual Studio Code.
2. Open the folder containing the cloned repository (genetic-programming-classification).

Step 3: Prepare the Dataset

Place your training and test datasets in the data directory (or any directory of your choice). Ensure the datasets are in CSV format.

Step 4: Run the Program

1. Open the Main.java file in VS Code.
2. Modify the file paths in the Main.java file to point to your training and test datasets:

   List<double[]> trainingData = loadData("data/mushroom_train.csv");
   List<double[]> testData = loadData("data/mushroom_test.csv");

3. Run the program by clicking the Run button in VS Code or by using the terminal:

   javac Main.java
   java Main

---

📊 Results

After running the program, you will see the following outputs:

1. Training Accuracy: The accuracy of the best individual on the training set for each generation.
2. Test Metrics: The accuracy, specificity, sensitivity, and F-measure of the best individual on the test set.

Example Output

Generation 0 Accuracy: 0.9998860509707512
Generation 1 Accuracy: 0.9998860509707512
...
Test Accuracy: 58.80398671096345%
Test Specificity: 0.0
Test Sensitivity: 1.0
Test F-measure: 0.7405857740585774

---

🤝 Contributing

Contributions are welcome! If you'd like to contribute, please follow these steps:

1. Fork the repository.
2. Create a new branch for your feature or bugfix.
3. Commit your changes.
4. Submit a pull request.

---

📜 License

This project is licensed under the MIT License. See the LICENSE file for details.

---

🙏 Acknowledgments

• Inspired by the field of Genetic Programming and evolutionary algorithms.
• Built using Java and Visual Studio Code.

---

🎉 Enjoy!

Feel free to explore, experiment, and contribute to this project. If you have any questions or suggestions, don't hesitate to open an issue or reach out!

---