Heart Attack Risk Prediction
Logistic regression and hypothesis testing to explore lifestyle factors associated with heart attack risk.
Project Snapshot
Analyzed a real-world health dataset to predict heart attack risk using logistic regression and hypothesis testing, focusing on individuals without family history. Explored associations between exercise, sleep, and stress levels through EDA, model building, and permutation tests in R.
Problem
Heart attacks remain a leading cause of death. Understanding which lifestyle variables — exercise, sleep, stress — are statistically associated with heart attack risk can inform preventive strategies, especially for individuals without genetic predisposition.
Approach
- Performed exploratory data analysis on health and lifestyle variables
- Built logistic regression models for binary heart attack risk prediction
- Conducted permutation-based hypothesis tests to assess statistical significance
- Focused analysis on the subset of individuals without family history of heart disease
Tech Stack
R, ggplot2, dplyr, tidyverse
Results
- None of the tested lifestyle variables showed statistically significant effects in hypothesis testing
- Prompted critical discussion on data limitations, sample size, and model interpretation
- Demonstrated rigorous application of statistical methodology even with null results