Breast cancer remains one of the most prevalent and life-threatening diseases among women worldwide. Early detection significantly improves survival rates, yet traditional diagnostic methods such as mammography and biopsy can be time-consuming and prone to human error. To address these challenges, computational intelligence techniques offer promising solutions to enhance accuracy and efficiency in breast cancer detection.

🔹 This project explores the application of Neural Networks (NN), Particle Swarm Optimization (PSO), Feedforward Neural Networks (FNN), and Convolutional Neural Networks (CNN) to improve breast cancer diagnosis.
🔹 By integrating advanced machine learning algorithms, we aim to develop a reliable and interpretable system for breast cancer classification.

Despite advancements in medical imaging and AI-based diagnostics, several challenges persist in breast cancer detection:

🔸 Diagnostic Errors – Traditional methods may lead to misdiagnosis due to subjective interpretation of medical images.
🔸 Limited Dataset Size – Small datasets can result in overfitting and poor generalization in deep learning models.
🔸 Feature Selection Complexity – Identifying the most relevant features from medical images and datasets remains a challenge.
🔸 Optimization Constraints – Traditional backpropagation methods may not efficiently optimize neural networks for medical diagnosis.

This project aims to:

✅ Improve breast cancer detection accuracy using computational intelligence techniques.
✅ Evaluate the performance of Neural Networks (NN), Particle Swarm Optimization (PSO), Feedforward Neural Networks (FNN), and Convolutional Neural Networks (CNN).
✅ Enhance feature selection and classification techniques to optimize predictive performance.
✅ Develop a scalable and robust diagnostic model that minimizes false positives and false negatives.

🚀 Neural Network + Particle Swarm Optimization (NN + PSO) – Utilizes PSO to optimize NN weights, improving classification accuracy.
🎯 Feedforward Neural Network (FNN) – Implements a multi-layer perceptron with backpropagation for pattern recognition.
🖼 Convolutional Neural Network (CNN) – Employs deep learning techniques for image-based breast cancer classification.
🔬 Dataset – Utilizes Wisconsin Breast Cancer Dataset (Kaggle), containing malignant and benign tumor records.
📊 Performance Metrics – Accuracy, Precision, Recall, and F1-Score are used for model evaluation.

The performance of each model was evaluated using accuracy, precision, recall, and F1-score:

💡 Key Insights:

• CNN achieved the highest accuracy (97.37%), making it the most effective model for breast cancer classification.
• FNN performed well but exhibited slight overfitting.
• NN + PSO showed lower performance, indicating the need for further optimization.

🔹 Data Augmentation – Enhance dataset diversity for better generalization.
🔹 Transfer Learning – Utilize pre-trained deep learning models for improved feature extraction.
🔹 Hyperparameter Optimization – Fine-tune model parameters for improved efficiency.
🔹 Ensemble Learning – Combine multiple models to improve classification accuracy.

We would like to express our gratitude to the following individuals for their contributions:

• Muhammad Ariff Ridzlan
• Siti Nur Aisyah
• Nurul Hurul Aini
• Siti Nabila