Stroke Prediction Project

Overview

This project predicts the likelihood of a stroke using machine learning and deep learning techniques. It includes exploratory data analysis (EDA), data preprocessing, and model training with traditional and ensemble learning methods.

Features and Methodology

Exploratory Data Analysis (EDA)

• Analyzing feature distributions
• Correlation analysis with stroke occurrence

Data Preprocessing

• Handling missing values
• Feature scaling and encoding
• SMOTE for dataset balancing

Machine Learning Models

• KNN, Logistic Regression, Decision Tree, Naive Bayes, SVM

Ensemble Learning Models

• Voting Classifier, Bagging Classifier, Stacking Classifier

Deep Learning Models

• Neural Networks (MLP) with dropout & early stopping
• Recurrent Neural Networks (RNN) with hyperparameter tuning

Libraries Used

Data Manipulation & Visualization

• pandas, numpy, seaborn, matplotlib

Machine Learning

• sklearn, imblearn

Deep Learning

• tensorflow, keras

How to Use

1. Clone the repository and install dependencies:

   git clone <repo_link>
   cd stroke_prediction
   pip install -r requirements.txt

2. Run the Jupyter Notebook for data analysis and model training.
3. Evaluate models using accuracy, confusion matrix, and classification reports.

Results

• Multiple models are evaluated to determine the best-performing classifier.
• Stacking Classifier and Neural Networks achieve the highest accuracy.