Abstract

Heart disease is one of the leading causes of death globally, and early prediction can significantly reduce fatal outcomes. This project presents a machine learning-based approach to predict the likelihood of a heart attack based on various health indicators such as age, gender, cholesterol level, blood pressure, and chest pain type. The model was trained and evaluated using real-world datasets and achieved promising results in classification accuracy.

Introduction

Cardiovascular diseases have become a major public health concern. Traditional diagnosis often requires invasive procedures, whereas machine learning techniques can provide fast, reliable, and cost-effective predictions. The goal of this research is to develop a predictive model that helps in identifying high-risk individuals using clinical data.

Methodology

1. Data Collection:
   The dataset was obtained from the UCI Machine Learning Repository (Heart Disease Dataset).

2. Data Preprocessing:

   • Handled missing values.
   • Normalized numerical attributes.
   • Encoded categorical features.

3. Model Selection:
   Multiple algorithms were tested, including:

   • Logistic Regression
   • Random Forest
   • Support Vector Machine (SVM)
   • K-Nearest Neighbors (KNN)

4. Model Training and Validation:

   • The dataset was split into 80% training and 20% testing sets.
   • Cross-validation was used to avoid overfitting.
   • Evaluation metrics included Accuracy, Precision, Recall, and F1-Score.

Experiments

Different models were trained and compared using Python (Scikit-learn). Hyperparameter tuning was performed using GridSearchCV to find optimal model settings.
Below is a sample code snippet:

from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier(n_estimators=100, random_state=42)
model.fit(X_train, y_train)
y_pred = model.predict(X_test)
<!-- RT_DIVIDER -->
# Results
The Random Forest model provided the best results with:

Accuracy: 89%

Precision: 88%

Recall: 90%

F1-Score: 89%
<!-- RT_DIVIDER -->
# Conclusion
This study demonstrates the potential of machine learning in predicting heart attack risk using basic clinical parameters. Future work can include integration with healthcare applications for real-time prediction and expanding the dataset for better generalization.