This project requires to design and implement a medical data classification system using three different approaches: Support Vector Machine (SVM), Convolutional Neural Network (CNN), and Particle Swarm Optimization (PSO). The goal is to perform medical data recognition and classification using standard medical datasets. must compare the performance of traditional Machine Learning (SVM), Deep Learning (CNN), and Optimization-based approaches (PSO) for classification tasks. The report should include configuration code snippets and an eNSP topology file, demonstrating a cohesive and collaborative effort by the group. Each group must use a STANDARD medical dataset from a reliable source such as: • Kaggle (e.g., Breast Cancer, Pneumonia X-ray, Diabetic Retinopathy) • UCI Machine Learning Repository (e.g., Breast Cancer Wisconsin Dataset) • NIH Medical Imaging datasets Dataset must contain clear classification labels. Dataset source must be cited properly in APA format. • Load and explore the dataset. • Perform preprocessing (normalization, encoding, resizing for images). • Split data into training, validation, and testing sets. • Show dataset statistics and visualizations. • Implement Support Vector Machine for classification. • Test different kernels (linear, RBF). • Tune hyperparameters. • Report performance metrics. • Design and train a Convolutional Neural Network. • Show architecture details (layers, activation functions, optimizer, loss function). • Plot training and validation curves. • Evaluate performance on test set. • Implement Particle Swarm Optimization. • Use PSO for hyperparameter optimization or feature selection. • Compare results before and after optimization. • must evaluate all models using: • Accuracy • Errors Measurements like: EMC,REMC • Precision • Recall • F1-score • Confusion Matrix • ROC Curve (if applicable) • Provide a comparison table of SVM vs CNN vs PSO-optimized model. Each group must demonstrate a recognition system where a new medical sample (image or data input) is classified using the trained model.
السؤال
Aaisha Al Buraiki
This project requires to design and implement a medical data classification system using three different approaches: Support Vector Machine (SVM), Convolutional Neural Network (CNN), and Particle Swarm Optimization (PSO). The goal is to perform medical data recognition and classification using standard medical datasets. must compare the performance of traditional Machine Learning (SVM), Deep Learning (CNN), and Optimization-based approaches (PSO) for classification tasks. The report should include configuration code snippets and an eNSP topology file, demonstrating a cohesive and collaborative effort by the group. Each group must use a STANDARD medical dataset from a reliable source such as: • Kaggle (e.g., Breast Cancer, Pneumonia X-ray, Diabetic Retinopathy) • UCI Machine Learning Repository (e.g., Breast Cancer Wisconsin Dataset) • NIH Medical Imaging datasets Dataset must contain clear classification labels. Dataset source must be cited properly in APA format. • Load and explore the dataset. • Perform preprocessing (normalization, encoding, resizing for images). • Split data into training, validation, and testing sets. • Show dataset statistics and visualizations. • Implement Support Vector Machine for classification. • Test different kernels (linear, RBF). • Tune hyperparameters. • Report performance metrics. • Design and train a Convolutional Neural Network. • Show architecture details (layers, activation functions, optimizer, loss function). • Plot training and validation curves. • Evaluate performance on test set. • Implement Particle Swarm Optimization. • Use PSO for hyperparameter optimization or feature selection. • Compare results before and after optimization. • must evaluate all models using: • Accuracy • Errors Measurements like: EMC,REMC • Precision • Recall • F1-score • Confusion Matrix • ROC Curve (if applicable) • Provide a comparison table of SVM vs CNN vs PSO-optimized model. Each group must demonstrate a recognition system where a new medical sample (image or data input) is classified using the trained model.
1 جواب على هذا السؤال
Recommended Posts
انضم إلى النقاش
يمكنك أن تنشر الآن وتسجل لاحقًا. إذا كان لديك حساب، فسجل الدخول الآن لتنشر باسم حسابك.