pearl franz

selected projects in machine learning and software

Heart Disease Diagnosis

Python: scikit-learn, seaborn, matplotlib

This project utilizes exploratory data analysis (EDA) and a Support Vector Machine (SVM) with an RBF kernel to predict the presence of heart disease based on clinical features. Through feature visualization and classification modeling, the system achieved an accuracy of 88.5% and an AUC of 0.94 on the test set.

Key insights:

Distributions of features like age, maximum heart rate achieved (thalach), sex, and chest pain type revealed clear separability between patients with and without heart disease.
The ROC curve demonstrates strong classification performance with minimal false positives.
The confusion matrix and evaluation metrics (precision = 0.89, recall = 0.86, F1 = 0.87) support the model’s reliability across both classes.

This work highlights the potential of interpretable machine learning models to assist with early and accurate diagnosis of cardiovascular conditions.

CMU Virtual Art Installation

Python: tkinter

This project is an interactive drawing application designed as a proof of concept for a collaborative digital art installation at Carnegie Mellon. Inspired by public art pieces that grow through community input, the app allows users to add colored dots to a shared canvas, imagining a future in which students from different disciplines contribute to evolving visual pieces tied to locations across campus.

The project was recognized with an Honorary Mention at HackCMU 2020, highlighting its creativity and potential for cross-disciplinary engagement.

Page updated

Google Sites

Report abuse