This project demonstrates an end-to-end statistical modeling workflow using a simulated prostate cancer dataset to evaluate how demographic, lifestyle, and clinical factors influence prostate cancer risk severity.
Using R and Quarto, I performed:
- Data cleaning and preprocessing
- Exploratory data analysis (EDA)
- Summary statistics and data visualization
- Multiple linear regression modeling
- Ordinal logistic regression modeling
- AIC-based model selection
- Model diagnostics and assumption testing
- Predicted probability visualization for risk interpretation
The project used approximately 1,000 patient records containing demographic, lifestyle, and clinical variables to identify predictors of prostate cancer severity and communicate findings through reproducible analysis.
This project highlights my experience working with structured healthcare datasets, building predictive models, and translating statistical outputs into actionable insights.
Tools Used: R, Quarto, ggplot2, MASS, dplyr
Files Included:
- PDF report with full analysis
- Quarto source code
- Dataset source linked below
Dataset Source:https://www.kaggle.com/datasets/miadul/prostate-cancer-risk-and-lifestyle-synthetic-dataset
Note: This project was originally completed as an academic statistical modeling project. The PDF reflects the original analysis environment, including a local file path reference, but the repository contains the full reproducible workflow and methodology.