karthikanchipaku/Effective-Hypertension-Detection-Using-predictive-Feature-Engineering-And-DeepLearning

# 🩺 基于预测性特征工程和深度学习的高效高血压检测 这是一款先进的医疗分类系统，结合了严密的特性工程与混合 Gradient Boosting 和 LSTM 深度学习架构，能够以极高的诊断准确率预测高血压。 ## 🏗️ 系统架构 本项目围绕一个稳健的临床预测 pipeline 构建： * **特征工程与预处理** – 高级数据清洗、特征选择，并使用 SMOTE 缓解类别不平衡问题，以处理倾斜的临床数据集。 * **混合建模引擎** – 将 gradient boosting 分类器与 Long Short-Term Memory (LSTM) 深度学习网络相结合，以同时捕获表格模式和序列指标。 * **推理层** – 轻量级的 Flask backend 提供实时预测，并配有一个交互式 Web 用户界面。 ## 🚀 技术栈 * **机器学习与深度学习：** Python, PyTorch, Scikit-learn, XGBoost, LSTM, SMOTE * **后端与 API：** Python, Flask * **数据处理与分析：** Pandas, NumPy, Matplotlib, Seaborn * **部署与版本控制：** Git, GitHub, Local Web Server ## 📁 仓库结构 ``` Effective-Hypertension-Detection-Using-predictive-Feature-Engineering-And-DeepLearning/ ├── data/ # Clinical datasets and preprocessing files ├── models/ # Saved PyTorch and XGBoost model weights ├── notebooks/ # Exploratory data analysis and model training ├── src/ # Core pipeline modules │ ├── preprocessing.py # Data cleaning, SMOTE, and scaling │ ├── feature_engineering.py # Feature selection and transformation │ └── train_hybrid.py # Hybrid model training scripts ├── app.py # Flask backend application runner ├── templates/ # HTML UI templates for the web interface ├── static/ # CSS stylesheets and static assets ├── requirements.txt # Python dependencies ├── .gitignore └── README.md 📈 Key Features High Diagnostic Accuracy – Achieves 94% prediction accuracy through optimized hybrid feature-model architectures. Advanced Class Imbalance Handling – Employs SMOTE (Synthetic Minority Over-sampling Technique) to ensure reliable classification across varied patient cohorts. Hybrid Deep Learning Framework – Combines tabular gradient boosting strengths with LSTM sequential data modeling. Interactive Web Interface – Real-time patient metric input form via Flask to instantly compute hypertension risk assessments. 💻 Local Setup & Installation Prerequisites Python 3.10+ Git 1. Clone the repository git clone [https://github.com/karthikanchipaku/Effective-Hypertension-Detection-Using-predictive-Feature-Engineering-And-DeepLearning.git](https://github.com/karthikanchipaku/Effective-Hypertension-Detection-Using-predictive-Feature-Engineering-And-DeepLearning.git) cd Effective-Hypertension-Detection-Using-predictive-Feature-Engineering-And-DeepLearning 2. Create & activate virtual environment python -m venv venv # Windows venv\Scripts\activate # macOS / Linux source venv/bin/activate 3. Install dependencies pip install -r requirements.txt 4. Run the Flask Web Application Since app.py is located in the root directory, start the application directly with: python app.py Web interface will be available at: http://127.0.0.1:5000 📡 Sample Application Routes MethodRouteDescriptionGET/Renders the primary patient health metric input UIPOST/predictProcesses clinical input features and returns hypertension risk status 🛠️ How It Works (High Level) Data Preprocessing – Patient vital statistics and health indicators are cleaned, normalized, and balanced using SMOTE. Feature Optimization – Key predictive indicators are engineered to maximize model sensitivity and specificity. Hybrid Classification – Features are passed through the combined XGBoost and LSTM pipeline for final binary classification. Web Delivery – The Flask server handles incoming requests from the frontend interface and displays immediate diagnostic probability outputs. 📄 License This project is open source and available under the MIT License. ```

标签：Apex, Flask, 凭据扫描, 医疗预测, 数据科学, 机器学习, 深度学习, 特征工程, 资源验证, 逆向工具